diff --git a/README.md b/README.md index b187bb7e7d837a367ccd0862441947ad412c77f7..5bc647a155f0834780b5363c001b0f3f7514ac47 100644 --- a/README.md +++ b/README.md @@ -1,3 +1,114 @@ --- -license: cc-by-4.0 +pretty_name: VIEW2SPACE v1 +task_categories: +- visual-question-answering +tags: +- multi-view +- vision-language +- spatial-reasoning +- counting +- detection +- mcq --- + +# VIEW2SPACE v1 + +VIEW2SPACE v1 is a multi-view vision-language evaluation dataset for spatial reasoning. + +The public release is organized into three subsets: + +- `count` +- `detect` +- `mcq` + +For official code, updates, and usage instructions, please see: + +https://github.com/pokerme7777/VIEW2SPACE + +## Directory Structure + +```text +view2space-v1-release/ + README.md + count/ + overall.jsonl + detect/ + overall.jsonl + mcq/ + overall.jsonl + images/ + img_000001.png + img_000002.png + ... +``` + +Each subset contains one `overall.jsonl` file. Each line is one question. + +## Data Format + +Each JSONL record contains the following public fields: + +- `q_idx`: question id, such as `mcq_000001` +- `q_type`: question family, one of `mcq`, `detect`, or `count` +- `question`: question text +- `options`: multiple-choice options for MCQ questions; empty for non-MCQ questions +- `question_prompt`: extra prompt text when applicable +- `answer`: ground-truth answer +- `image_paths`: relative paths to the images used by the question +- `supporting.draw_boxes`: optional input boxes associated with one or more images + +Notes: + +- `image_paths` are relative to the dataset root, not to the subset folder. +- Example image path: `images/img_000123.png` +- `q_type` uses only the public family labels `mcq`, `detect`, and `count`. + +## Example Record + +```json +{ + "q_idx": "mcq_000001", + "q_type": "mcq", + "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, consider the nearest 'man with an orange hat, crouching'. Where is the 'white temple' relative to this 'man with an orange hat, crouching'?", + "options": { + "A": "left", + "B": "right", + "C": "front", + "D": "back" + }, + "question_prompt": "", + "answer": "C", + "image_paths": [ + "images/img_000813.png", + "images/img_000806.png", + "images/img_000815.png" + ], + "supporting": { + "draw_boxes": null + } +} +``` + +## Basic Usage + +```python +import json +from pathlib import Path + +dataset_root = Path("/path/to/view2space-v1-release") +jsonl_path = dataset_root / "mcq" / "overall.jsonl" + +with jsonl_path.open("r", encoding="utf-8") as f: + first = json.loads(next(f)) + +image_files = [dataset_root / rel_path for rel_path in first["image_paths"]] +print(first["q_idx"]) +print(first["q_type"]) +print(image_files) +``` + +## Citation + +If you use VIEW2SPACE v1, please refer to the official repository for the latest citation information: + +https://github.com/pokerme7777/VIEW2SPACE diff --git a/count/overall.jsonl b/count/overall.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..c4238b4aad90cee0db05093f37470d03c9c76bbe --- /dev/null +++ b/count/overall.jsonl @@ -0,0 +1,591 @@ +{"q_idx": "count_000001", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many woman are present in this scene?", "options": {}, "question_prompt": "", "answer": 9, "image_paths": ["images/img_001073.png", "images/img_001070.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000002", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many F-22 fighter jet with gray color are present in this scene?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000958.png", "images/img_000962.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000003", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many stack of bricks are present in this scene?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000445.png", "images/img_000446.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000004", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many helicopter are present in this scene?", "options": {}, "question_prompt": "", "answer": 10, "image_paths": ["images/img_001020.png", "images/img_001025.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000005", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many man are present in this scene?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_000733.png", "images/img_000737.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000006", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many man are present in this scene?", "options": {}, "question_prompt": "", "answer": 10, "image_paths": ["images/img_000069.png", "images/img_000066.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000007", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many blue painted metal bench without armrests are present in this scene?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000620.png", "images/img_000622.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000008", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many helicopter are present in this scene?", "options": {}, "question_prompt": "", "answer": 9, "image_paths": ["images/img_000991.png", "images/img_000988.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000009", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many worker wearing face mask and red shirt is present in this scene?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000479.png", "images/img_000472.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000010", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many civil airplane are present in this scene?", "options": {}, "question_prompt": "", "answer": 10, "image_paths": ["images/img_000262.png", "images/img_000258.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000011", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many basketball are present in this scene?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_000586.png", "images/img_000589.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000012", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many hot air balloon are present in this scene?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000959.png", "images/img_000957.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000013", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many man are present in this scene?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000877.png", "images/img_000876.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000014", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many airplane are present in this scene?", "options": {}, "question_prompt": "", "answer": 9, "image_paths": ["images/img_000212.png", "images/img_000211.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000015", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many drone with angular black body are present in this scene?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000359.png", "images/img_000358.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000016", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many beach cabin with striped blue color is present in this scene?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000673.png", "images/img_000670.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000017", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many tombstone with cross are present in this scene?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_001198.png", "images/img_001201.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000018", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many basketball are present in this scene?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000584.png", "images/img_000591.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000019", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many hedge sculpture with car-like shape is present in this scene?", "options": {}, "question_prompt": "", "answer": 0, "image_paths": ["images/img_000166.png", "images/img_000161.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000020", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many fighter jet are present in this scene?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_000896.png", "images/img_000898.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000021", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many basketball are present in this scene?", "options": {}, "question_prompt": "", "answer": 9, "image_paths": ["images/img_000572.png", "images/img_000569.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000022", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many truck with blue and white cab are present in this scene?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000512.png", "images/img_000509.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000023", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many fighter jet are present in this scene?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000274.png", "images/img_000269.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000024", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many tombstone with shoulder top is present in this scene?", "options": {}, "question_prompt": "", "answer": 0, "image_paths": ["images/img_000863.png", "images/img_000867.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000025", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many fighter jet are present in this scene?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000298.png", "images/img_000305.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000026", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many green shipping container are present in this scene?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000126.png", "images/img_000133.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000027", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many tombstone with cross are present in this scene?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000884.png", "images/img_000880.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000028", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many F-16 fighter jet are present in this scene?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000954.png", "images/img_000962.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000029", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many warehouse with pink roof is present in this scene?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000528.png", "images/img_000526.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000030", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many swim ring are present in this scene?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000703.png", "images/img_000702.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000031", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many green shipping container are present in this scene?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000093.png", "images/img_000094.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000032", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many shipping container are present in this scene?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_000081.png", "images/img_000084.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000033", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many helicopter are present in this scene?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_001008.png", "images/img_001004.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000034", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many tombstone without cross is present in this scene?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000833.png", "images/img_000830.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000035", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many dark grey helicopter are present in this scene?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_001044.png", "images/img_001043.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000036", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many truck are present in this scene?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000526.png", "images/img_000524.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000037", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many man are present in this scene?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_001139.png", "images/img_001133.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000038", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many harrier jumpjet fighter jet are present in this scene?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000332.png", "images/img_000334.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000039", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many helicopter are present in this scene?", "options": {}, "question_prompt": "", "answer": 10, "image_paths": ["images/img_001042.png", "images/img_001036.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000040", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many basketball are present in this scene?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000574.png", "images/img_000567.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000041", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many airplane with yellow livery are present in this scene?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000261.png", "images/img_000258.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000042", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many person are present in this scene?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000688.png", "images/img_000682.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000043", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many airplane with white and green livery is present in this scene?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000245.png", "images/img_000240.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000044", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many shipping container are present in this scene?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_000084.png", "images/img_000086.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000045", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many child are present in this scene?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000620.png", "images/img_000622.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000046", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many man are present in this scene?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_000012.png", "images/img_000014.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000047", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many beige sand helicopter are present in this scene?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000975.png", "images/img_000974.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000048", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many man are present in this scene?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000880.png", "images/img_000882.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000049", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many man are present in this scene?", "options": {}, "question_prompt": "", "answer": 9, "image_paths": ["images/img_000024.png", "images/img_000021.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000050", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many bench are present in this scene?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000624.png", "images/img_000621.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000051", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many fighter jet are present in this scene?", "options": {}, "question_prompt": "", "answer": 10, "image_paths": ["images/img_000955.png", "images/img_000954.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000052", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many stick arms snowman are present in this scene?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000734.png", "images/img_000736.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000053", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many airplane with red livery are present in this scene?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000204.png", "images/img_000201.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000054", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many basketball are present in this scene?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000603.png", "images/img_000606.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000055", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many drone with white oval body shape are present in this scene?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000400.png", "images/img_000397.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000056", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many tombstone with cross are present in this scene?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_001154.png", "images/img_001155.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000057", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many man are present in this scene?", "options": {}, "question_prompt": "", "answer": 9, "image_paths": ["images/img_000770.png", "images/img_000766.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000058", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many white truck with flatbed are present in this scene?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000539.png", "images/img_000541.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000059", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many snowman with stick arms and black buttons are present in this scene?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000730.png", "images/img_000729.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000060", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many black and orange lamp are present in this scene?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000476.png", "images/img_000479.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000061", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many shipping container are present in this scene?", "options": {}, "question_prompt": "", "answer": 10, "image_paths": ["images/img_000130.png", "images/img_000132.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000062", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many man wearing gray windbreaker are present in this scene?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000021.png", "images/img_000029.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000063", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many beige tombstone are present in this scene?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_001166.png", "images/img_001171.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000064", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many tombstone without cross is present in this scene?", "options": {}, "question_prompt": "", "answer": 0, "image_paths": ["images/img_000867.png", "images/img_000863.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000065", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many hot air balloon are present in this scene?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_001042.png", "images/img_001039.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000066", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many basketball are present in this scene?", "options": {}, "question_prompt": "", "answer": 9, "image_paths": ["images/img_000569.png", "images/img_000570.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000067", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many worker are present in this scene?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000428.png", "images/img_000422.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000068", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many upright tombstone are present in this scene?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_001132.png", "images/img_001136.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000069", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many drone with rectangular white and black body is present in this scene?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000348.png", "images/img_000346.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000070", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many fighter jet are present in this scene?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_000894.png", "images/img_000897.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000071", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many street lamp are present in this scene?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000152.png", "images/img_000149.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000072", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many bench with armrests are present in this scene?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000569.png", "images/img_000572.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000073", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many bench made of natural wood are present in this scene?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000601.png", "images/img_000607.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000074", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many woman are present in this scene?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_001057.png", "images/img_001051.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000075", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many street lamp are present in this scene?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000152.png", "images/img_000155.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000076", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many pink flamingo swim ring is present in this scene?", "options": {}, "question_prompt": "", "answer": 0, "image_paths": ["images/img_000705.png", "images/img_000697.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000077", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many fighter jet are present in this scene?", "options": {}, "question_prompt": "", "answer": 9, "image_paths": ["images/img_000909.png", "images/img_000907.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000078", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many drone are present in this scene?", "options": {}, "question_prompt": "", "answer": 16, "image_paths": ["images/img_000380.png", "images/img_000383.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000079", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many watering can is present in this scene?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000148.png", "images/img_000152.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000080", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many street lamp are present in this scene?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000154.png", "images/img_000148.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000081", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many stack of bricks are present in this scene?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000428.png", "images/img_000423.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000082", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many fighter jet are present in this scene?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000270.png", "images/img_000274.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000083", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many drone with white oval body shape are present in this scene?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000394.png", "images/img_000397.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000084", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many person wearing blue patterned shirt is present in this scene?", "options": {}, "question_prompt": "", "answer": 0, "image_paths": ["images/img_000655.png", "images/img_000649.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000085", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many food stall is present in this scene?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_001104.png", "images/img_001108.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000086", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many worker wearing face mask and red shirt is present in this scene?", "options": {}, "question_prompt": "", "answer": 0, "image_paths": ["images/img_000477.png", "images/img_000478.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000087", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many gray trolley are present in this scene?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000456.png", "images/img_000460.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000088", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many fighter jet are present in this scene?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000287.png", "images/img_000286.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000089", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many shipping container are present in this scene?", "options": {}, "question_prompt": "", "answer": 10, "image_paths": ["images/img_000495.png", "images/img_000493.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000090", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many hedge sculpture are present in this scene?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000153.png", "images/img_000149.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000091", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many swim ring are present in this scene?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000697.png", "images/img_000698.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000092", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many upright tombstone are present in this scene?", "options": {}, "question_prompt": "", "answer": 9, "image_paths": ["images/img_001204.png", "images/img_001202.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000093", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many surfboard with red and white design are present in this scene?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000671.png", "images/img_000665.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000094", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many black upright round tombstone is present in this scene?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000879.png", "images/img_000880.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000095", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many man are present in this scene?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000015.png", "images/img_000008.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000096", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many tombstone are present in this scene?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000833.png", "images/img_000834.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000097", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many snowman with black hat and black buttons are present in this scene?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000768.png", "images/img_000763.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000098", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many shipping container are present in this scene?", "options": {}, "question_prompt": "", "answer": 9, "image_paths": ["images/img_000091.png", "images/img_000093.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000099", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many upright tombstone are present in this scene?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_001152.png", "images/img_001151.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000100", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many woman are present in this scene?", "options": {}, "question_prompt": "", "answer": 10, "image_paths": ["images/img_001103.png", "images/img_001104.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000101", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many tombstone are present in this scene?", "options": {}, "question_prompt": "", "answer": 10, "image_paths": ["images/img_001197.png", "images/img_001200.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000102", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many snowman are present in this scene?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000798.png", "images/img_000795.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000103", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many woman in white clothing standing are present in this scene?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_001074.png", "images/img_001072.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000104", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many umbrella are present in this scene?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000687.png", "images/img_000684.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000105", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many warehouse with pitched roof are present in this scene?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000556.png", "images/img_000555.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000106", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many weathered metal bus shelter with side walls is present in this scene?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000007.png", "images/img_000012.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000107", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many drone with white oval body shape are present in this scene?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000399.png", "images/img_000398.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000108", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many red truck is present in this scene?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000456.png", "images/img_000463.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000109", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many concrete mixer are present in this scene?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000441.png", "images/img_000442.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000110", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many ball are present in this scene?", "options": {}, "question_prompt": "", "answer": 25, "image_paths": ["images/img_000587.png", "images/img_000584.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000111", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many street lamp are present in this scene?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000188.png", "images/img_000189.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000112", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many helicopter are present in this scene?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_000971.png", "images/img_000974.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000113", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many snowman with black hat, red scarf, and black buttons are present in this scene?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000747.png", "images/img_000751.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000114", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many man are present in this scene?", "options": {}, "question_prompt": "", "answer": 10, "image_paths": ["images/img_000786.png", "images/img_000780.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000115", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many airplane with yellow livery are present in this scene?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000260.png", "images/img_000259.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000116", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many small house are present in this scene?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000187.png", "images/img_000182.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000117", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many shipping container are present in this scene?", "options": {}, "question_prompt": "", "answer": 9, "image_paths": ["images/img_000106.png", "images/img_000107.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000118", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many hot air balloon with pink and maroon stripes is present in this scene?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000961.png", "images/img_000960.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000119", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many civil airplane are present in this scene?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_000204.png", "images/img_000196.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000120", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many civil airplane are present in this scene?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_000200.png", "images/img_000201.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000121", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many bench are present in this scene?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000603.png", "images/img_000608.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000122", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many helicopter are present in this scene?", "options": {}, "question_prompt": "", "answer": 9, "image_paths": ["images/img_001009.png", "images/img_001011.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000123", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many woman in beige clothing are present in this scene?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_001084.png", "images/img_001088.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000124", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many tombstone are present in this scene?", "options": {}, "question_prompt": "", "answer": 10, "image_paths": ["images/img_001183.png", "images/img_001187.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000125", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many ball are present in this scene?", "options": {}, "question_prompt": "", "answer": 26, "image_paths": ["images/img_000617.png", "images/img_000625.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000126", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many truck are present in this scene?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000540.png", "images/img_000544.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000127", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many woman are present in this scene?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_001051.png", "images/img_001058.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000128", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many fighter jet are present in this scene?", "options": {}, "question_prompt": "", "answer": 10, "image_paths": ["images/img_000941.png", "images/img_000943.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000129", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many man wearing a bandana and a denim vest, sitting is present in this scene?", "options": {}, "question_prompt": "", "answer": 0, "image_paths": ["images/img_000811.png", "images/img_000819.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000130", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many tombstone are present in this scene?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000876.png", "images/img_000882.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000131", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many white tombstone are present in this scene?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_001188.png", "images/img_001183.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000132", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many snowman are present in this scene?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000729.png", "images/img_000734.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000133", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many man are present in this scene?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_000733.png", "images/img_000736.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000134", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many drone with white oval body shape are present in this scene?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000346.png", "images/img_000344.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000135", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many shipping container are present in this scene?", "options": {}, "question_prompt": "", "answer": 11, "image_paths": ["images/img_000551.png", "images/img_000556.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000136", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many green street lamp are present in this scene?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000154.png", "images/img_000156.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000137", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many snowman with two balls and stick arms are present in this scene?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000786.png", "images/img_000785.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000138", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many airplane with red livery are present in this scene?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000242.png", "images/img_000245.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000139", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many small house with concrete material is present in this scene?", "options": {}, "question_prompt": "", "answer": 0, "image_paths": ["images/img_000175.png", "images/img_000171.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000140", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many snowman with three balls, red scarf, and black buttons is present in this scene?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000737.png", "images/img_000734.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000141", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many drone with angular black body are present in this scene?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000367.png", "images/img_000372.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000142", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many woman wearing scarf accessory are present in this scene?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_001083.png", "images/img_001088.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000143", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many airplane with light blue livery are present in this scene?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000318.png", "images/img_000312.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000144", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many woman are present in this scene?", "options": {}, "question_prompt": "", "answer": 9, "image_paths": ["images/img_001091.png", "images/img_001089.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000145", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many stick arms snowman are present in this scene?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000763.png", "images/img_000770.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000146", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many trellis are present in this scene?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000176.png", "images/img_000175.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000147", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many truck are present in this scene?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000555.png", "images/img_000551.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000148", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many shipping container are present in this scene?", "options": {}, "question_prompt": "", "answer": 10, "image_paths": ["images/img_000120.png", "images/img_000117.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000149", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many upright tombstone are present in this scene?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_001164.png", "images/img_001168.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000150", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many baseball is present in this scene?", "options": {}, "question_prompt": "", "answer": 0, "image_paths": ["images/img_000642.png", "images/img_000635.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000151", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many man are present in this scene?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_000015.png", "images/img_000007.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000152", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many blue shipping container are present in this scene?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000106.png", "images/img_000109.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000153", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many fighter jet are present in this scene?", "options": {}, "question_prompt": "", "answer": 9, "image_paths": ["images/img_000927.png", "images/img_000922.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000154", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many truck are present in this scene?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000510.png", "images/img_000506.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000155", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many woman wearing scarf accessory are present in this scene?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_001089.png", "images/img_001091.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000156", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many airplane are present in this scene?", "options": {}, "question_prompt": "", "answer": 9, "image_paths": ["images/img_000230.png", "images/img_000227.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000157", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many basketball are present in this scene?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000591.png", "images/img_000586.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000158", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many trolley is present in this scene?", "options": {}, "question_prompt": "", "answer": 0, "image_paths": ["images/img_000479.png", "images/img_000478.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000159", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many man are present in this scene?", "options": {}, "question_prompt": "", "answer": 9, "image_paths": ["images/img_000041.png", "images/img_000042.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000160", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many tombstone are present in this scene?", "options": {}, "question_prompt": "", "answer": 9, "image_paths": ["images/img_001168.png", "images/img_001171.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000161", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many drone are present in this scene?", "options": {}, "question_prompt": "", "answer": 16, "image_paths": ["images/img_000394.png", "images/img_000393.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000162", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many drone with white oval body shape are present in this scene?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000350.png", "images/img_000348.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000163", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many man wearing red jacket are present in this scene?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000073.png", "images/img_000069.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000164", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many small house with gray rood and concrete material is present in this scene?", "options": {}, "question_prompt": "", "answer": 0, "image_paths": ["images/img_000141.png", "images/img_000144.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000165", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many fighter jet are present in this scene?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_000318.png", "images/img_000315.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000166", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many woman wearing high boots are present in this scene?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_001084.png", "images/img_001089.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000167", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many person with black hat are present in this scene?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000652.png", "images/img_000655.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000168", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many olive green helicopter is present in this scene?", "options": {}, "question_prompt": "", "answer": 0, "image_paths": ["images/img_001011.png", "images/img_001010.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000169", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many lamp is present in this scene?", "options": {}, "question_prompt": "", "answer": 0, "image_paths": ["images/img_000444.png", "images/img_000445.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000170", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many tombstone without cross are present in this scene?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_001184.png", "images/img_001187.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000171", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many drone with rectangular white and black body are present in this scene?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000381.png", "images/img_000386.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000172", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many flat tombstone are present in this scene?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000834.png", "images/img_000830.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000173", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many warehouse with pitched roof are present in this scene?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000079.png", "images/img_000084.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000174", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many airplane with white and green livery are present in this scene?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000202.png", "images/img_000203.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000175", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many truck are present in this scene?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000557.png", "images/img_000551.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000176", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many orange warehouse with pitched roof is present in this scene?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000078.png", "images/img_000083.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000177", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many airplane with yellow livery is present in this scene?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000335.png", "images/img_000333.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000178", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many man are present in this scene?", "options": {}, "question_prompt": "", "answer": 9, "image_paths": ["images/img_000753.png", "images/img_000749.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000179", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many drone with smooth black curved body are present in this scene?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000371.png", "images/img_000368.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000180", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many black tombstone are present in this scene?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_001148.png", "images/img_001154.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000181", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many street lamp are present in this scene?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000184.png", "images/img_000187.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000182", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many F-22 fighter jet with gray color are present in this scene?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000317.png", "images/img_000319.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000183", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many black and orange lamp are present in this scene?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000480.png", "images/img_000473.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000184", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many beige upright shoulder tombstone without cross is present in this scene?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_001205.png", "images/img_001200.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000185", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many red painted metal bench with armrests is present in this scene?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000591.png", "images/img_000583.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000186", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many drone with white oval body shape are present in this scene?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000370.png", "images/img_000373.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000187", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many bus stop with enclosure are present in this scene?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000015.png", "images/img_000010.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000188", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many street lamp are present in this scene?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000182.png", "images/img_000189.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000189", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many fighter jet are present in this scene?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000302.png", "images/img_000305.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000190", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many white and green striped hot air balloon is present in this scene?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_001028.png", "images/img_001020.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000191", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many drone with white oval body shape are present in this scene?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000350.png", "images/img_000346.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000192", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000580.png", "images/img_000589.png", "images/img_000583.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000193", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 3 but not in view 1?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_000560.png", "images/img_000566.png", "images/img_000570.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000194", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'tombstone with pointed top' are visible in view 3 but not in view 1?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_001174.png", "images/img_001187.png", "images/img_001181.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000195", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'man wearing gray windbreaker' are visible in view 1 but not in view 3?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000023.png", "images/img_000028.png", "images/img_000017.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000196", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 2 but not in view 3?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_000573.png", "images/img_000570.png", "images/img_000560.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000197", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'drone with white oval body shape' are visible in view 1 but not in view 2?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000349.png", "images/img_000342.png", "images/img_000348.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000198", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'green street lamp' are visible in view 1 but not in view 3?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000149.png", "images/img_000148.png", "images/img_000145.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000199", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'blue shipping container' are visible in view 1 but not in view 2?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000493.png", "images/img_000484.png", "images/img_000496.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000200", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'airplane with red livery' are visible in view 3 but not in view 2?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000240.png", "images/img_000236.png", "images/img_000243.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000201", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'F-16 fighter jet' are visible in view 1 but not in view 3?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000912.png", "images/img_000909.png", "images/img_000905.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000202", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'beach ball' are visible in view 1 but not in view 3?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000621.png", "images/img_000619.png", "images/img_000614.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000203", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 3 but not in view 2?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000641.png", "images/img_000627.png", "images/img_000636.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000204", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'man wearing black coat' are visible in view 3 but not in view 1?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000004.png", "images/img_000009.png", "images/img_000013.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000205", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'drone with white oval body shape' are visible in view 1 but not in view 3?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000344.png", "images/img_000345.png", "images/img_000338.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000206", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'drone with white oval body shape' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000391.png", "images/img_000398.png", "images/img_000395.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000207", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 1 but not in view 3?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_000570.png", "images/img_000568.png", "images/img_000560.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000208", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 1 but not in view 3?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000590.png", "images/img_000587.png", "images/img_000580.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000209", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'lamp' are visible in view 3 but not in view 2?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000475.png", "images/img_000467.png", "images/img_000473.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000210", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'man wearing santa costume' are visible in view 3 but not in view 2?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000799.png", "images/img_000791.png", "images/img_000798.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000211", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'man wearing gray windbreaker' are visible in view 3 but not in view 1?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000020.png", "images/img_000024.png", "images/img_000028.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000212", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'man wearing santa costume' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000064.png", "images/img_000070.png", "images/img_000066.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000213", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'beach ball' are visible in view 3 but not in view 1?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000599.png", "images/img_000603.png", "images/img_000607.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000214", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 3 but not in view 1?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000580.png", "images/img_000590.png", "images/img_000587.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000215", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 3 but not in view 1?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000580.png", "images/img_000586.png", "images/img_000589.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000216", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'F-22 fighter jet with gray color' are visible in view 1 but not in view 3?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000315.png", "images/img_000319.png", "images/img_000307.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000217", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'drone with white oval body shape' are visible in view 3 but not in view 1?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000388.png", "images/img_000394.png", "images/img_000396.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000218", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'drone with white oval body shape' are visible in view 2 but not in view 3?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000396.png", "images/img_000395.png", "images/img_000388.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000219", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 1 but not in view 2?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_000570.png", "images/img_000560.png", "images/img_000566.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000220", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'black street lamp' are visible in view 1 but not in view 2?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000160.png", "images/img_000157.png", "images/img_000161.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000221", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 1 but not in view 3?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000587.png", "images/img_000588.png", "images/img_000581.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000222", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'drone with white oval body shape' are visible in view 1 but not in view 2?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000393.png", "images/img_000388.png", "images/img_000394.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000223", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'drone with white oval body shape' are visible in view 2 but not in view 3?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000347.png", "images/img_000350.png", "images/img_000340.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000224", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'snowman with red gloves and black buttons' are visible in view 1 but not in view 2?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000748.png", "images/img_000739.png", "images/img_000746.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000225", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'baseball' are visible in view 3 but not in view 2?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000623.png", "images/img_000611.png", "images/img_000619.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000226", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 3 but not in view 1?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000580.png", "images/img_000587.png", "images/img_000589.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000227", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'beach ball' are visible in view 1 but not in view 2?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000624.png", "images/img_000616.png", "images/img_000620.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000228", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 3 but not in view 2?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000585.png", "images/img_000580.png", "images/img_000587.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000229", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 1 but not in view 3?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000590.png", "images/img_000586.png", "images/img_000580.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000230", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'red shipping container' are visible in view 1 but not in view 3?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000524.png", "images/img_000527.png", "images/img_000515.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000231", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'drone with white oval body shape' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000342.png", "images/img_000350.png", "images/img_000349.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000232", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'stylized F22 fighter jet with grey and cyan color' are visible in view 3 but not in view 2?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000961.png", "images/img_000946.png", "images/img_000956.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000233", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000559.png", "images/img_000568.png", "images/img_000570.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000234", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 2 but not in view 3?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000589.png", "images/img_000590.png", "images/img_000580.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000235", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'green camouflage helicopter' are visible in view 3 but not in view 1?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_001033.png", "images/img_001036.png", "images/img_001040.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000236", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 2 but not in view 3?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_000571.png", "images/img_000566.png", "images/img_000559.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000237", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'trolley' are visible in view 2 but not in view 3?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000462.png", "images/img_000455.png", "images/img_000447.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000238", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 3 but not in view 1?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000580.png", "images/img_000588.png", "images/img_000589.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000239", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 3 but not in view 1?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000580.png", "images/img_000583.png", "images/img_000587.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000240", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'airplane with white and orange livery' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000277.png", "images/img_000284.png", "images/img_000288.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000241", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'drone with rectangular white and black body' are visible in view 1 but not in view 3?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000360.png", "images/img_000358.png", "images/img_000354.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000242", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'F-22 fighter jet with gray color' are visible in view 1 but not in view 3?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000316.png", "images/img_000314.png", "images/img_000309.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000243", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 3 but not in view 2?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000587.png", "images/img_000580.png", "images/img_000590.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000244", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000580.png", "images/img_000587.png", "images/img_000586.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000245", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'drone with white oval body shape' are visible in view 1 but not in view 3?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000346.png", "images/img_000343.png", "images/img_000337.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000246", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 3 but not in view 2?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000567.png", "images/img_000563.png", "images/img_000573.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000247", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'drone with white oval body shape' are visible in view 1 but not in view 3?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000349.png", "images/img_000345.png", "images/img_000342.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000248", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_000560.png", "images/img_000570.png", "images/img_000572.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000249", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'blue shipping container' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000101.png", "images/img_000108.png", "images/img_000106.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000250", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 1 but not in view 3?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_000566.png", "images/img_000569.png", "images/img_000559.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000251", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'drone with white oval body shape' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000389.png", "images/img_000398.png", "images/img_000394.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000252", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'drone with white oval body shape' are visible in view 2 but not in view 3?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000358.png", "images/img_000361.png", "images/img_000353.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000253", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'drone with white oval body shape' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000337.png", "images/img_000346.png", "images/img_000344.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000254", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'mirage fighter jet' are visible in view 3 but not in view 2?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000938.png", "images/img_000936.png", "images/img_000942.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000255", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'dark grey helicopter' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000969.png", "images/img_000978.png", "images/img_000972.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000256", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'drone with white oval body shape' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000387.png", "images/img_000395.png", "images/img_000397.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000257", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 1 but not in view 2?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000589.png", "images/img_000580.png", "images/img_000588.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000258", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'baseball' are visible in view 2 but not in view 3?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000584.png", "images/img_000589.png", "images/img_000581.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000259", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 2 but not in view 3?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_000572.png", "images/img_000566.png", "images/img_000559.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000260", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'drone with white oval body shape' are visible in view 1 but not in view 2?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000346.png", "images/img_000337.png", "images/img_000349.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000261", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'drone with white oval body shape' are visible in view 2 but not in view 3?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000346.png", "images/img_000350.png", "images/img_000341.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000262", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000598.png", "images/img_000607.png", "images/img_000603.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000263", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000580.png", "images/img_000586.png", "images/img_000584.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000264", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'tennis ball' are visible in view 3 but not in view 2?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000567.png", "images/img_000565.png", "images/img_000571.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000265", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 3 but not in view 2?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000585.png", "images/img_000580.png", "images/img_000589.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000266", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'blue swim ring' are visible in view 3 but not in view 1?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000690.png", "images/img_000701.png", "images/img_000699.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000267", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 1 but not in view 2?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_000570.png", "images/img_000560.png", "images/img_000568.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000268", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'man wearing red jacket' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000035.png", "images/img_000042.png", "images/img_000038.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000269", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'tennis ball' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000593.png", "images/img_000603.png", "images/img_000600.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000270", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'green shipping container' are visible in view 1 but not in view 2?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000084.png", "images/img_000077.png", "images/img_000083.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000271", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'drone with white oval body shape' are visible in view 3 but not in view 2?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000346.png", "images/img_000337.png", "images/img_000345.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000272", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'tennis ball' are visible in view 3 but not in view 2?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000569.png", "images/img_000560.png", "images/img_000566.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000273", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'beach ball' are visible in view 2 but not in view 3?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000606.png", "images/img_000604.png", "images/img_000597.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000274", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 1 but not in view 2?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000589.png", "images/img_000580.png", "images/img_000590.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000275", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 3 but not in view 1?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_000559.png", "images/img_000568.png", "images/img_000566.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000276", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'airplane with red livery' are visible in view 2 but not in view 3?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000269.png", "images/img_000270.png", "images/img_000265.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000277", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'airplane with yellow livery' are visible in view 3 but not in view 1?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000253.png", "images/img_000256.png", "images/img_000259.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000278", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'man wearing a green hoodie and orange shoes' are visible in view 2 but not in view 3?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000861.png", "images/img_000864.png", "images/img_000858.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000279", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'stylized F22 fighter jet with grey and cyan color' are visible in view 2 but not in view 3?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000959.png", "images/img_000961.png", "images/img_000951.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000280", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'concrete mixer' are visible in view 3 but not in view 2?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000440.png", "images/img_000437.png", "images/img_000443.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000281", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'drone with white oval body shape' are visible in view 1 but not in view 3?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000346.png", "images/img_000347.png", "images/img_000340.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000282", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'airplane with yellow livery' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000250.png", "images/img_000257.png", "images/img_000259.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000283", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'red shipping container' are visible in view 3 but not in view 1?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000110.png", "images/img_000118.png", "images/img_000114.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000284", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'man wearing gray windbreaker' are visible in view 2 but not in view 3?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000786.png", "images/img_000780.png", "images/img_000771.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000285", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'man wearing a green hoodie and orange shoes' are visible in view 3 but not in view 2?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000860.png", "images/img_000852.png", "images/img_000859.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000286", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'harrier jumpjet fighter jet' are visible in view 2 but not in view 3?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000945.png", "images/img_000941.png", "images/img_000935.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000287", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'drone with white oval body shape' are visible in view 3 but not in view 1?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000337.png", "images/img_000350.png", "images/img_000343.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000288", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'tennis ball' are visible in view 3 but not in view 1?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000560.png", "images/img_000573.png", "images/img_000566.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000289", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'trolley' are visible in view 1 but not in view 3?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000409.png", "images/img_000413.png", "images/img_000402.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000290", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'man wearing gray hat' are visible in view 3 but not in view 1?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000726.png", "images/img_000731.png", "images/img_000734.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000291", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'beach ball' are visible in view 3 but not in view 1?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000595.png", "images/img_000605.png", "images/img_000604.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000292", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'baseball' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000579.png", "images/img_000587.png", "images/img_000585.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000293", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'beach ball' are visible in view 1 but not in view 2?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000622.png", "images/img_000616.png", "images/img_000623.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000294", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 1 but not in view 2?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_000566.png", "images/img_000559.png", "images/img_000573.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000295", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000581.png", "images/img_000590.png", "images/img_000583.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000296", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'beach ball' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000660.png", "images/img_000668.png", "images/img_000665.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000297", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'drone with white oval body shape' are visible in view 2 but not in view 3?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000347.png", "images/img_000348.png", "images/img_000339.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000298", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000580.png", "images/img_000587.png", "images/img_000585.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000299", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'white shipping container' are visible in view 1 but not in view 3?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000490.png", "images/img_000495.png", "images/img_000484.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000300", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'drone with white oval body shape' are visible in view 3 but not in view 2?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000350.png", "images/img_000342.png", "images/img_000349.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000301", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 2 but not in view 3?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_000567.png", "images/img_000570.png", "images/img_000560.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000302", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'red shipping container' are visible in view 3 but not in view 2?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000522.png", "images/img_000515.png", "images/img_000520.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000303", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'lamp' are visible in view 1 but not in view 3?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000474.png", "images/img_000478.png", "images/img_000467.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000304", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'tombstone with pointed top' are visible in view 3 but not in view 1?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_001172.png", "images/img_001185.png", "images/img_001182.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000305", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 2 but not in view 3?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000600.png", "images/img_000607.png", "images/img_000599.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000306", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'man wearing brown coat' are visible in view 1 but not in view 3?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000040.png", "images/img_000037.png", "images/img_000031.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000307", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'green shipping container' are visible in view 1 but not in view 2?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000541.png", "images/img_000535.png", "images/img_000538.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000308", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 2 but not in view 3?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000583.png", "images/img_000590.png", "images/img_000580.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000309", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'black street lamp' are visible in view 1 but not in view 3?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000142.png", "images/img_000141.png", "images/img_000136.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000310", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'tombstone with shoulder top' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_001173.png", "images/img_001182.png", "images/img_001186.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000311", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'soccer ball' are visible in view 1 but not in view 3?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000573.png", "images/img_000566.png", "images/img_000565.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000312", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'airplane with yellow livery' are visible in view 3 but not in view 2?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000259.png", "images/img_000251.png", "images/img_000257.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000313", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'man in crouching posture' are visible in view 1 but not in view 3?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000830.png", "images/img_000831.png", "images/img_000824.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000314", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'beach ball' are visible in view 2 but not in view 3?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000634.png", "images/img_000641.png", "images/img_000631.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000315", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'baseball' are visible in view 3 but not in view 2?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000583.png", "images/img_000580.png", "images/img_000589.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000316", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'drone with white oval body shape' are visible in view 1 but not in view 3?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000393.png", "images/img_000398.png", "images/img_000387.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000317", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'tombstone with pointed top' are visible in view 3 but not in view 2?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_001183.png", "images/img_001172.png", "images/img_001180.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000318", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'airplane with yellow livery' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000250.png", "images/img_000258.png", "images/img_000254.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000319", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 2 but not in view 3?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_000568.png", "images/img_000566.png", "images/img_000559.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000320", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'red shipping container' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000514.png", "images/img_000520.png", "images/img_000521.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000321", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'truck with blue and white cab' are visible in view 3 but not in view 2?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000509.png", "images/img_000497.png", "images/img_000507.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000322", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 3 but not in view 2?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_000573.png", "images/img_000560.png", "images/img_000566.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000323", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000580.png", "images/img_000590.png", "images/img_000584.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000324", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'blue shipping container' are visible in view 3 but not in view 2?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000507.png", "images/img_000504.png", "images/img_000511.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000325", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'drone with white oval body shape' are visible in view 3 but not in view 2?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000349.png", "images/img_000339.png", "images/img_000348.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000326", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'dark grey helicopter' are visible in view 1 but not in view 2?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_001010.png", "images/img_001000.png", "images/img_001004.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000327", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'trolley' are visible in view 3 but not in view 1?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000450.png", "images/img_000457.png", "images/img_000456.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000328", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'drone with white oval body shape' are visible in view 3 but not in view 2?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000399.png", "images/img_000387.png", "images/img_000394.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000329", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'drone with white oval body shape' are visible in view 3 but not in view 2?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000356.png", "images/img_000352.png", "images/img_000360.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000330", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'trolley' are visible in view 1 but not in view 3?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000456.png", "images/img_000458.png", "images/img_000447.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000331", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'red shipping container' are visible in view 3 but not in view 1?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000517.png", "images/img_000524.png", "images/img_000522.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000332", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'airplane with yellow livery' are visible in view 1 but not in view 2?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000260.png", "images/img_000253.png", "images/img_000256.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000333", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'drone with white oval body shape' are visible in view 2 but not in view 3?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000347.png", "images/img_000350.png", "images/img_000342.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000334", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'drone with white oval body shape' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000337.png", "images/img_000343.png", "images/img_000348.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000335", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 3 but not in view 2?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_000567.png", "images/img_000559.png", "images/img_000566.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000336", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'lamp' are visible in view 1 but not in view 3?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000474.png", "images/img_000475.png", "images/img_000467.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000337", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 2 but not in view 3?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000590.png", "images/img_000589.png", "images/img_000581.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000338", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'drone with smooth black curved body' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000339.png", "images/img_000348.png", "images/img_000343.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000339", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'airplane with white and green livery' are visible in view 1 but not in view 2?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000197.png", "images/img_000192.png", "images/img_000196.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000340", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'tennis ball' are visible in view 3 but not in view 2?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000617.png", "images/img_000611.png", "images/img_000621.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000341", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'green street lamp' are visible in view 3 but not in view 2?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000183.png", "images/img_000178.png", "images/img_000182.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000342", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 3 but not in view 1?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_000560.png", "images/img_000567.png", "images/img_000570.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000343", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'drone with rectangular white and black body' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000375.png", "images/img_000382.png", "images/img_000385.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000344", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'drone with white oval body shape' are visible in view 2 but not in view 3?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000343.png", "images/img_000345.png", "images/img_000336.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000345", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_000560.png", "images/img_000566.png", "images/img_000573.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000346", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'woman in white clothing' are visible in view 3 but not in view 1?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_001045.png", "images/img_001051.png", "images/img_001053.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000347", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'airplane with light blue livery' are visible in view 1 but not in view 2?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000203.png", "images/img_000195.png", "images/img_000202.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000348", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_000560.png", "images/img_000566.png", "images/img_000572.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000349", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 3 but not in view 2?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000571.png", "images/img_000559.png", "images/img_000569.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000350", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'person saluting' are visible in view 3 but not in view 2?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000687.png", "images/img_000676.png", "images/img_000683.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000351", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'red shipping container' are visible in view 1 but not in view 2?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000526.png", "images/img_000517.png", "images/img_000523.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000352", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'bench made of white painted wood' are visible in view 3 but not in view 1?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000593.png", "images/img_000604.png", "images/img_000600.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000353", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'black flat tombstone with cross' are visible in view 1 but not in view 2?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_001165.png", "images/img_001158.png", "images/img_001166.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000354", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'woman in white clothing' are visible in view 2 but not in view 3?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_001051.png", "images/img_001057.png", "images/img_001049.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000355", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'woman in pink clothing' are visible in view 2 but not in view 3?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_001106.png", "images/img_001103.png", "images/img_001096.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000356", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'blue swim ring' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000689.png", "images/img_000699.png", "images/img_000698.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000357", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'blue shipping container' are visible in view 3 but not in view 1?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000124.png", "images/img_000130.png", "images/img_000132.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000358", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'drone with white oval body shape' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000341.png", "images/img_000350.png", "images/img_000344.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000359", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 3 but not in view 2?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_000572.png", "images/img_000560.png", "images/img_000566.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000360", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000580.png", "images/img_000587.png", "images/img_000590.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000361", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'beige sand helicopter' are visible in view 3 but not in view 2?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000977.png", "images/img_000966.png", "images/img_000972.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000362", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 1 but not in view 2?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000589.png", "images/img_000580.png", "images/img_000587.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000363", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'F-16 fighter jet' are visible in view 1 but not in view 3?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000287.png", "images/img_000288.png", "images/img_000278.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000364", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'blue shipping container' are visible in view 1 but not in view 3?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000130.png", "images/img_000132.png", "images/img_000123.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000365", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 2 but not in view 3?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_000571.png", "images/img_000566.png", "images/img_000560.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000366", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'man wearing gray windbreaker' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000772.png", "images/img_000782.png", "images/img_000785.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000367", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 1 but not in view 3?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000589.png", "images/img_000584.png", "images/img_000580.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000368", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'airplane with yellow livery' are visible in view 3 but not in view 1?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000252.png", "images/img_000260.png", "images/img_000261.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000369", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'beach ball' are visible in view 2 but not in view 3?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000637.png", "images/img_000639.png", "images/img_000633.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000370", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_000559.png", "images/img_000566.png", "images/img_000570.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000371", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 1 but not in view 2?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000589.png", "images/img_000580.png", "images/img_000584.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000372", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'surfboard' are visible in view 3 but not in view 2?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000671.png", "images/img_000662.png", "images/img_000670.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000373", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'beach ball' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000598.png", "images/img_000606.png", "images/img_000600.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000374", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 3 but not in view 2?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000589.png", "images/img_000580.png", "images/img_000590.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000375", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_000559.png", "images/img_000566.png", "images/img_000573.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000376", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 3 but not in view 1?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_000560.png", "images/img_000573.png", "images/img_000570.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000377", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000616.png", "images/img_000622.png", "images/img_000617.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000378", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'drone with white oval body shape' are visible in view 3 but not in view 2?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000348.png", "images/img_000338.png", "images/img_000346.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000379", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 1 but not in view 3?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_000566.png", "images/img_000572.png", "images/img_000559.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000380", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'drone with white oval body shape' are visible in view 1 but not in view 2?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000374.png", "images/img_000366.png", "images/img_000369.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000381", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'lamp' are visible in view 1 but not in view 2?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000473.png", "images/img_000467.png", "images/img_000474.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000382", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'drone with white oval body shape' are visible in view 3 but not in view 2?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000347.png", "images/img_000342.png", "images/img_000348.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000383", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 3 but not in view 1?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000595.png", "images/img_000604.png", "images/img_000605.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000384", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'drone with white oval body shape' are visible in view 1 but not in view 3?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000349.png", "images/img_000348.png", "images/img_000342.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000385", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'lamp' are visible in view 3 but not in view 1?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000467.png", "images/img_000475.png", "images/img_000476.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000386", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000580.png", "images/img_000589.png", "images/img_000584.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000387", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'bench made of black metal' are visible in view 2 but not in view 1?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000632.png", "images/img_000640.png", "images/img_000641.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000388", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'beach ball' are visible in view 1 but not in view 3?", "options": {}, "question_prompt": "", "answer": 6, "image_paths": ["images/img_000623.png", "images/img_000618.png", "images/img_000613.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000389", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 3 but not in view 1?", "options": {}, "question_prompt": "", "answer": 8, "image_paths": ["images/img_000560.png", "images/img_000567.png", "images/img_000566.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000390", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'blue swim ring' are visible in view 1 but not in view 3?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000703.png", "images/img_000697.png", "images/img_000696.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000391", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nHow many 'basketball' are visible in view 1 but not in view 3?", "options": {}, "question_prompt": "", "answer": 7, "image_paths": ["images/img_000590.png", "images/img_000588.png", "images/img_000580.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "count_000392", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('bench made of black metal').\nFor every instance, it has 1 basketball to its right-behind in view 2.\nIn view 1, how many 'basketball' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000570.png", "images/img_000560.png"], "supporting": {"draw_boxes": {"images/img_000560.png": []}}} +{"q_idx": "count_000393", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('bench made of black metal').\nFor every instance, it has 2 basketballs to its right-front, 4 basketballs to its left-front, and 1 bench made of black metal to its left in view 3.\nIn view 2, how many 'basketball' that lie in the left-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000573.png", "images/img_000571.png", "images/img_000564.png"], "supporting": {"draw_boxes": {"images/img_000564.png": []}}} +{"q_idx": "count_000394", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('stylized F22 fighter jet with grey and cyan color').\nFor every instance, it has 1 airplane with red livery to its right-behind in view 2.\nIn view 3, how many 'airplane with red livery' that lie in the right-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000270.png", "images/img_000266.png", "images/img_000276.png"], "supporting": {"draw_boxes": {"images/img_000266.png": []}}} +{"q_idx": "count_000395", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('person saluting').\nFor every instance, it has 1 person saluting and 1 person with black hat to its front in view 1.\nIn view 2, how many 'person saluting' that lie in the direct-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000695.png", "images/img_000703.png", "images/img_000701.png"], "supporting": {"draw_boxes": {"images/img_000695.png": []}}} +{"q_idx": "count_000396", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('blue painted metal bench without armrests').\nFor every instance, it has 2 basketballs to its right-behind, and 1 blue painted metal bench without armrests to its right in view 2.\nIn view 1, how many 'basketball' that lie in the right-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000617.png", "images/img_000611.png"], "supporting": {"draw_boxes": {"images/img_000611.png": []}}} +{"q_idx": "count_000397", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('bench made of natural wood').\nFor every instance, it has 5 basketballs to its left-behind, and 1 basketball to its right-behind in view 2.\nIn view 1, how many 'basketball' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000573.png", "images/img_000564.png"], "supporting": {"draw_boxes": {"images/img_000564.png": []}}} +{"q_idx": "count_000398", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('bench made of natural wood').\nFor every instance, it has 5 basketballs to its left-behind, and 1 basketball to its right-behind in view 3.\nIn view 2, how many 'basketball' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000570.png", "images/img_000573.png", "images/img_000564.png"], "supporting": {"draw_boxes": {"images/img_000564.png": []}}} +{"q_idx": "count_000399", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('person').\nFor every instance, it has 1 surfboard to its left in view 3.\nIn view 2, how many 'surfboard' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000669.png", "images/img_000670.png", "images/img_000660.png"], "supporting": {"draw_boxes": {"images/img_000660.png": []}}} +{"q_idx": "count_000400", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('worker wearing red hat and white shirt').\nFor every instance, it has 3 lamps to its left in view 1.\nIn view 3, how many 'lamp' that lie in the direct-left direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000466.png", "images/img_000476.png", "images/img_000472.png"], "supporting": {"draw_boxes": {"images/img_000466.png": []}}} +{"q_idx": "count_000401", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('man in crouching posture').\nFor every instance, it has 1 man in sitting posture and 1 man in crouching posture to its behind in view 1.\nIn view 2, how many 'man in sitting posture' that lie in the direct-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000823.png", "images/img_000832.png", "images/img_000827.png"], "supporting": {"draw_boxes": {"images/img_000823.png": []}}} +{"q_idx": "count_000402", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('natural wooden bench with armrests').\nFor every instance, it has 5 basketballs to its left-behind, and 1 basketball to its right-behind in view 2.\nIn view 1, how many 'basketball' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000573.png", "images/img_000564.png"], "supporting": {"draw_boxes": {"images/img_000564.png": []}}} +{"q_idx": "count_000403", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('mirage fighter jet').\nFor every instance, it has 1 airplane with white and orange livery to its right-front in view 1.\nIn view 3, how many 'airplane with white and orange livery' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000296.png", "images/img_000301.png", "images/img_000303.png"], "supporting": {"draw_boxes": {"images/img_000296.png": []}}} +{"q_idx": "count_000404", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('bench made of red painted metal').\nFor every instance, it has 2 basketballs to its left-front in view 3.\nIn view 2, how many 'basketball' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000624.png", "images/img_000621.png", "images/img_000615.png"], "supporting": {"draw_boxes": {"images/img_000615.png": []}}} +{"q_idx": "count_000405", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('bench made of natural wood').\nFor every instance, it has 5 basketballs to its left-behind, and 1 basketball to its right-behind in view 2.\nIn view 3, how many 'basketball' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000567.png", "images/img_000564.png", "images/img_000573.png"], "supporting": {"draw_boxes": {"images/img_000564.png": []}}} +{"q_idx": "count_000406", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('bench made of black metal').\nFor every instance, it has 2 basketballs to its right-behind, and 2 bench made of black metals to its right in view 1.\nIn view 3, how many 'basketball' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000632.png", "images/img_000640.png", "images/img_000641.png"], "supporting": {"draw_boxes": {"images/img_000632.png": []}}} +{"q_idx": "count_000407", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('green wooden bench with armrests').\nFor every instance, it has 2 basketballs to its left-behind in view 2.\nIn view 1, how many 'basketball' that lie in the left-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000622.png", "images/img_000615.png", "images/img_000624.png"], "supporting": {"draw_boxes": {"images/img_000615.png": []}}} +{"q_idx": "count_000408", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('bench made of black metal').\nFor every instance, it has 1 basketball to its left-behind in view 1.\nIn view 2, how many 'basketball' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000628.png", "images/img_000637.png"], "supporting": {"draw_boxes": {"images/img_000628.png": []}}} +{"q_idx": "count_000409", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('bench made of blue painted metal').\nFor every instance, it has 1 tennis ball to its left-front, and 1 blue painted metal bench without armrests to its right in view 1.\nIn view 2, how many 'tennis ball' that lie in the right-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000611.png", "images/img_000620.png"], "supporting": {"draw_boxes": {"images/img_000611.png": []}}} +{"q_idx": "count_000410", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('man wearing green hood').\nFor every instance, it has 1 man wearing red jacket to its right in view 1.\nIn view 2, how many 'man wearing green hood' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000058.png", "images/img_000066.png"], "supporting": {"draw_boxes": {"images/img_000058.png": []}}} +{"q_idx": "count_000411", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('airplane with red livery').\nFor every instance, it has 1 harrier jumpjet fighter jet to its right-behind in view 3.\nIn view 1, how many 'harrier jumpjet fighter jet' that lie in the right-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000329.png", "images/img_000330.png", "images/img_000320.png"], "supporting": {"draw_boxes": {"images/img_000320.png": []}}} +{"q_idx": "count_000412", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('bench made of natural wood').\nFor every instance, it has 3 basketballs to its right-front, 1 basketball to its front, and 1 bench made of natural wood to its right in view 1.\nIn view 3, how many 'basketball' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000598.png", "images/img_000606.png", "images/img_000605.png"], "supporting": {"draw_boxes": {"images/img_000598.png": []}}} +{"q_idx": "count_000413", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('lamp').\nFor every instance, it has 1 lamp to its left, 1 lamp to its right, and 1 forklift to its right-behind in view 1.\nIn view 3, how many 'lamp' that lie in the direct-right direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000465.png", "images/img_000474.png", "images/img_000475.png"], "supporting": {"draw_boxes": {"images/img_000465.png": []}}} +{"q_idx": "count_000414", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('F-16 fighter jet').\nFor every instance, it has 1 stylized F22 fighter jet with grey and cyan color to its behind in view 2.\nIn view 1, how many 'stylized F22 fighter jet with grey and cyan color' that lie in the right-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000960.png", "images/img_000950.png"], "supporting": {"draw_boxes": {"images/img_000950.png": []}}} +{"q_idx": "count_000415", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('stylized F22 fighter jet with grey and cyan color').\nFor every instance, it has 1 airplane with red livery to its right-behind in view 1.\nIn view 3, how many 'airplane with red livery' that lie in the right-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000267.png", "images/img_000269.png", "images/img_000276.png"], "supporting": {"draw_boxes": {"images/img_000267.png": []}}} +{"q_idx": "count_000416", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('green wooden bench with armrests').\nFor every instance, it has 1 basketball to its left-behind in view 1.\nIn view 3, how many 'basketball' that lie in the right-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000560.png", "images/img_000571.png", "images/img_000570.png"], "supporting": {"draw_boxes": {"images/img_000560.png": []}}} +{"q_idx": "count_000417", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('white upright tombstone with cross').\nFor every instance, it has 1 tombstone with pointed top to its left in view 2.\nIn view 1, how many 'tombstone with pointed top' that lie in the direct-left direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_001197.png", "images/img_001189.png"], "supporting": {"draw_boxes": {"images/img_001189.png": []}}} +{"q_idx": "count_000418", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('man wearing gray hat').\nFor every instance, it has 1 man wearing gray hat and 1 man wearing red jacket to its behind in view 1.\nIn view 2, how many 'man wearing red jacket' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000727.png", "images/img_000734.png", "images/img_000731.png"], "supporting": {"draw_boxes": {"images/img_000727.png": []}}} +{"q_idx": "count_000419", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('worker').\nFor every instance, it has 1 worker to its front, and 1 forklift to its right-behind in view 1.\nIn view 2, how many 'forklift' that lie in the right-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000436.png", "images/img_000443.png"], "supporting": {"draw_boxes": {"images/img_000436.png": []}}} +{"q_idx": "count_000420", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('bench made of natural wood').\nFor every instance, it has 5 basketballs to its left-behind, and 1 basketball to its right-behind in view 2.\nIn view 1, how many 'basketball' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000572.png", "images/img_000564.png"], "supporting": {"draw_boxes": {"images/img_000564.png": []}}} +{"q_idx": "count_000421", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('bench made of natural wood').\nFor every instance, it has 5 basketballs to its left-behind, and 1 basketball to its right-behind in view 3.\nIn view 2, how many 'basketball' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000569.png", "images/img_000570.png", "images/img_000564.png"], "supporting": {"draw_boxes": {"images/img_000564.png": []}}} +{"q_idx": "count_000422", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('red gloves snowman').\nFor every instance, it has 1 red gloves snowman to its right, and 1 SUV to its right-behind in view 2.\nIn view 1, how many 'red gloves snowman' that lie in the direct-right direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000783.png", "images/img_000776.png", "images/img_000782.png"], "supporting": {"draw_boxes": {"images/img_000776.png": []}}} +{"q_idx": "count_000423", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('F-16 fighter jet').\nFor every instance, it has 1 harrier jumpjet fighter jet to its behind in view 1.\nIn view 2, how many 'harrier jumpjet fighter jet' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000930.png", "images/img_000940.png"], "supporting": {"draw_boxes": {"images/img_000930.png": []}}} +{"q_idx": "count_000424", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('bench made of black metal').\nFor every instance, it has 1 tennis ball to its right-behind, 1 tennis ball to its behind, and 2 bench made of black metals to its right in view 1.\nIn view 3, how many 'tennis ball' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000632.png", "images/img_000635.png", "images/img_000641.png"], "supporting": {"draw_boxes": {"images/img_000632.png": []}}} +{"q_idx": "count_000425", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('red painted metal bench with armrests').\nFor every instance, it has 1 basketball to its right-behind, and 1 basketball to its behind in view 1.\nIn view 3, how many 'basketball' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000581.png", "images/img_000588.png", "images/img_000587.png"], "supporting": {"draw_boxes": {"images/img_000581.png": []}}} +{"q_idx": "count_000426", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('bench made of natural wood').\nFor every instance, it has 1 child in gray clothing to its right-front in view 1.\nIn view 2, how many 'bench made of natural wood' that lie in the direct-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000594.png", "images/img_000600.png"], "supporting": {"draw_boxes": {"images/img_000594.png": []}}} +{"q_idx": "count_000427", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('bench made of stone').\nFor every instance, it has 1 baseball to its right-front in view 2.\nIn view 1, how many 'baseball' that lie in the right-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000583.png", "images/img_000577.png", "images/img_000590.png"], "supporting": {"draw_boxes": {"images/img_000577.png": []}}} +{"q_idx": "count_000428", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('bench made of natural wood').\nFor every instance, it has 1 beach ball to its left-front, 1 beach ball to its front, and 1 bench made of natural wood to its left in view 1.\nIn view 2, how many 'beach ball' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000598.png", "images/img_000606.png"], "supporting": {"draw_boxes": {"images/img_000598.png": []}}} +{"q_idx": "count_000429", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('bench made of black metal').\nFor every instance, it has 2 basketballs to its right-front, 4 basketballs to its left-front, and 1 bench made of black metal to its left in view 2.\nIn view 1, how many 'basketball' that lie in the left-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000572.png", "images/img_000564.png"], "supporting": {"draw_boxes": {"images/img_000564.png": []}}} +{"q_idx": "count_000430", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('man wearing brown coat').\nFor every instance, it has 2 man wearing red jackets to its front in view 2.\nIn view 1, how many 'man wearing red jacket' that lie in the direct-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000013.png", "images/img_000005.png", "images/img_000010.png"], "supporting": {"draw_boxes": {"images/img_000005.png": []}}} +{"q_idx": "count_000431", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('bench made of natural wood').\nFor every instance, it has 1 beach ball to its left-front, 1 beach ball to its front, and 1 natural wooden bench with armrests to its left in view 2.\nIn view 1, how many 'beach ball' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000605.png", "images/img_000598.png"], "supporting": {"draw_boxes": {"images/img_000598.png": []}}} +{"q_idx": "count_000432", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('toy crab').\nFor every instance, it has 1 beach ball to its left in view 3.\nIn view 1, how many 'beach ball' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000718.png", "images/img_000715.png", "images/img_000712.png"], "supporting": {"draw_boxes": {"images/img_000712.png": []}}} +{"q_idx": "count_000433", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('natural wooden bench with armrests').\nFor every instance, it has 5 basketballs to its left-behind, and 1 basketball to its right-behind in view 1.\nIn view 2, how many 'basketball' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000564.png", "images/img_000572.png", "images/img_000570.png"], "supporting": {"draw_boxes": {"images/img_000564.png": []}}} +{"q_idx": "count_000434", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('fan').\nFor every instance, it has 2 workers to its right in view 1.\nIn view 3, how many 'worker' that lie in the direct-right direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000453.png", "images/img_000462.png", "images/img_000459.png"], "supporting": {"draw_boxes": {"images/img_000453.png": []}}} +{"q_idx": "count_000435", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('natural wooden bench with armrests').\nFor every instance, it has 1 beach ball to its left-front, 1 beach ball to its front, and 1 natural wooden bench with armrests to its left in view 2.\nIn view 1, how many 'beach ball' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000607.png", "images/img_000598.png"], "supporting": {"draw_boxes": {"images/img_000598.png": []}}} +{"q_idx": "count_000436", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('bench made of black metal').\nFor every instance, it has 2 basketballs to its right-behind, and 2 bench made of black metals to its right in view 3.\nIn view 1, how many 'basketball' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000641.png", "images/img_000639.png", "images/img_000632.png"], "supporting": {"draw_boxes": {"images/img_000632.png": []}}} +{"q_idx": "count_000437", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('woman wearing black hat').\nFor every instance, it has 1 woman wearing black hat to its left, and 1 food stall with milk branding to its left-front in view 2.\nIn view 1, how many 'woman wearing black hat' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_001083.png", "images/img_001077.png"], "supporting": {"draw_boxes": {"images/img_001077.png": []}}} +{"q_idx": "count_000438", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('bench made of natural wood').\nFor every instance, it has 1 beach ball to its left-front, 1 beach ball to its front, and 1 bench made of natural wood to its left in view 2.\nIn view 1, how many 'beach ball' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000605.png", "images/img_000598.png"], "supporting": {"draw_boxes": {"images/img_000598.png": []}}} +{"q_idx": "count_000439", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('man wearing brown coat').\nFor every instance, it has 1 snowman with black hat and stick arms to its left-front in view 3.\nIn view 2, how many 'snowman with black hat and stick arms' that lie in the left-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000749.png", "images/img_000748.png", "images/img_000739.png"], "supporting": {"draw_boxes": {"images/img_000739.png": []}}} +{"q_idx": "count_000440", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('F-16 fighter jet').\nFor every instance, it has 1 harrier jumpjet fighter jet to its behind in view 3.\nIn view 1, how many 'harrier jumpjet fighter jet' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000940.png", "images/img_000942.png", "images/img_000930.png"], "supporting": {"draw_boxes": {"images/img_000930.png": []}}} +{"q_idx": "count_000441", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('fan').\nFor every instance, it has 2 workers to its right in view 3.\nIn view 2, how many 'worker' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000461.png", "images/img_000460.png", "images/img_000453.png"], "supporting": {"draw_boxes": {"images/img_000453.png": []}}} +{"q_idx": "count_000442", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('man wearing green hood').\nFor every instance, it has 1 man wearing gray windbreaker to its behind in view 1.\nIn view 2, how many 'man wearing gray windbreaker' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000020.png", "images/img_000026.png", "images/img_000024.png"], "supporting": {"draw_boxes": {"images/img_000020.png": []}}} +{"q_idx": "count_000443", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('beige sand helicopter').\nFor every instance, it has 2 beige sand helicopters and 1 olive green helicopter to its right in view 1.\nIn view 2, how many 'beige sand helicopter' that lie in the direct-right direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000965.png", "images/img_000972.png"], "supporting": {"draw_boxes": {"images/img_000965.png": []}}} +{"q_idx": "count_000444", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('F-16 fighter jet').\nFor every instance, it has 1 stylized F22 fighter jet with grey and cyan color to its behind in view 2.\nIn view 1, how many 'stylized F22 fighter jet with grey and cyan color' that lie in the right-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000957.png", "images/img_000950.png"], "supporting": {"draw_boxes": {"images/img_000950.png": []}}} +{"q_idx": "count_000445", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('worker wearing red hat and white shirt').\nFor every instance, it has 3 lamps to its left in view 1.\nIn view 3, how many 'lamp' that lie in the direct-left direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000466.png", "images/img_000479.png", "images/img_000473.png"], "supporting": {"draw_boxes": {"images/img_000466.png": []}}} +{"q_idx": "count_000446", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('red curved metal bench without armrests').\nFor every instance, it has 2 basketballs to its right-front in view 1.\nIn view 2, how many 'basketball' that lie in the right-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000581.png", "images/img_000589.png"], "supporting": {"draw_boxes": {"images/img_000581.png": []}}} +{"q_idx": "count_000447", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('desert camouflage helicopter').\nFor every instance, it has 2 olive green helicopters to its behind in view 1.\nIn view 2, how many 'olive green helicopter' that lie in the direct-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_001012.png", "images/img_001022.png", "images/img_001025.png"], "supporting": {"draw_boxes": {"images/img_001012.png": []}}} +{"q_idx": "count_000448", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('man wearing gray windbreaker').\nFor every instance, it has 2 man wearing gray windbreakers to its right, and 1 police car to its right-front in view 1.\nIn view 3, how many 'man wearing gray windbreaker' that lie in the direct-right direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000774.png", "images/img_000785.png", "images/img_000783.png"], "supporting": {"draw_boxes": {"images/img_000774.png": []}}} +{"q_idx": "count_000449", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('bench made of blue painted metal').\nFor every instance, it has 2 basketballs to its right-front, and 1 bench made of blue painted metal to its right in view 2.\nIn view 1, how many 'basketball' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000617.png", "images/img_000611.png"], "supporting": {"draw_boxes": {"images/img_000611.png": []}}} +{"q_idx": "count_000450", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('dark grey helicopter').\nFor every instance, it has 1 dark grey helicopter to its front, and 1 white and green striped hot air balloon to its right-front in view 1.\nIn view 2, how many 'white and green striped hot air balloon' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000963.png", "images/img_000971.png"], "supporting": {"draw_boxes": {"images/img_000963.png": []}}} +{"q_idx": "count_000451", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('stylized F22 fighter jet with grey and cyan color').\nFor every instance, it has 1 stylized F22 fighter jet with grey and cyan color to its right, and 1 green and black striped hot air balloon to its left-behind in view 3.\nIn view 2, how many 'stylized F22 fighter jet with grey and cyan color' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000954.png", "images/img_000959.png", "images/img_000953.png"], "supporting": {"draw_boxes": {"images/img_000953.png": []}}} +{"q_idx": "count_000452", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('red painted metal bench with armrests').\nFor every instance, it has 2 soccer balls to its right-behind in view 2.\nIn view 3, how many 'soccer ball' that lie in the right-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000590.png", "images/img_000581.png", "images/img_000589.png"], "supporting": {"draw_boxes": {"images/img_000581.png": []}}} +{"q_idx": "count_000453", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('man wearing brown coat').\nFor every instance, it has 1 man wearing santa costume to its front in view 3.\nIn view 1, how many 'man wearing santa costume' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000071.png", "images/img_000068.png", "images/img_000064.png"], "supporting": {"draw_boxes": {"images/img_000064.png": []}}} +{"q_idx": "count_000454", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('worker wearing red hat and white shirt').\nFor every instance, it has 3 lamps to its left in view 3.\nIn view 2, how many 'lamp' that lie in the direct-left direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000474.png", "images/img_000475.png", "images/img_000466.png"], "supporting": {"draw_boxes": {"images/img_000466.png": []}}} +{"q_idx": "count_000455", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('man wearing gray hat').\nFor every instance, it has 2 red gloves snowmans to its left-behind in view 2.\nIn view 1, how many 'red gloves snowman' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000784.png", "images/img_000777.png", "images/img_000785.png"], "supporting": {"draw_boxes": {"images/img_000777.png": []}}} +{"q_idx": "count_000456", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('airplane with white and orange livery').\nFor every instance, it has 1 airplane with white and green livery to its behind in view 3.\nIn view 1, how many 'airplane with white and green livery' that lie in the right-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000230.png", "images/img_000228.png", "images/img_000221.png"], "supporting": {"draw_boxes": {"images/img_000221.png": []}}} +{"q_idx": "count_000457", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('mirage fighter jet').\nFor every instance, it has 1 airplane with red livery to its left-front in view 2.\nIn view 1, how many 'airplane with red livery' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000270.png", "images/img_000265.png"], "supporting": {"draw_boxes": {"images/img_000265.png": []}}} +{"q_idx": "count_000458", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('bench made of natural wood').\nFor every instance, it has 4 baseballs to its left-front, and 1 bench made of natural wood to its left in view 2.\nIn view 1, how many 'baseball' that lie in the left-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000605.png", "images/img_000598.png"], "supporting": {"draw_boxes": {"images/img_000598.png": []}}} +{"q_idx": "count_000459", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('natural wooden bench with armrests').\nFor every instance, it has 5 beach balls to its right-front in view 3.\nIn view 1, how many 'beach ball' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000600.png", "images/img_000603.png", "images/img_000594.png"], "supporting": {"draw_boxes": {"images/img_000594.png": []}}} +{"q_idx": "count_000460", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('natural wooden bench with armrests').\nFor every instance, it has 5 beach balls to its right-front in view 1.\nIn view 2, how many 'beach ball' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000594.png", "images/img_000604.png", "images/img_000603.png"], "supporting": {"draw_boxes": {"images/img_000594.png": []}}} +{"q_idx": "count_000461", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('bench made of natural wood').\nFor every instance, it has 5 basketballs to its left-behind, and 1 basketball to its right-behind in view 2.\nIn view 1, how many 'basketball' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000571.png", "images/img_000564.png", "images/img_000572.png"], "supporting": {"draw_boxes": {"images/img_000564.png": []}}} +{"q_idx": "count_000462", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('natural wooden bench with armrests').\nFor every instance, it has 1 beach ball to its left-front, 1 beach ball to its front, and 1 natural wooden bench with armrests to its left in view 1.\nIn view 2, how many 'beach ball' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000598.png", "images/img_000606.png"], "supporting": {"draw_boxes": {"images/img_000598.png": []}}} +{"q_idx": "count_000463", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('olive green helicopter').\nFor every instance, it has 1 red and white helicopter to its front in view 2.\nIn view 3, how many 'olive green helicopter' that lie in the left-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_001039.png", "images/img_001029.png", "images/img_001037.png"], "supporting": {"draw_boxes": {"images/img_001029.png": []}}} +{"q_idx": "count_000464", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('man wearing santa costume').\nFor every instance, it has 1 red gloves snowman to its right-behind in view 1.\nIn view 2, how many 'red gloves snowman' that lie in the right-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000778.png", "images/img_000786.png"], "supporting": {"draw_boxes": {"images/img_000778.png": []}}} +{"q_idx": "count_000465", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('bench made of black metal').\nFor every instance, it has 2 basketballs to its right-front, 4 basketballs to its left-front, and 1 bench made of black metal to its left in view 3.\nIn view 1, how many 'basketball' that lie in the left-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000571.png", "images/img_000566.png", "images/img_000564.png"], "supporting": {"draw_boxes": {"images/img_000564.png": []}}} +{"q_idx": "count_000466", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('stylized F22 fighter jet with white and red color').\nFor every instance, it has 1 airplane with white and orange livery to its right-behind in view 1.\nIn view 2, how many 'airplane with white and orange livery' that lie in the right-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000280.png", "images/img_000287.png", "images/img_000285.png"], "supporting": {"draw_boxes": {"images/img_000280.png": []}}} +{"q_idx": "count_000467", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('F-22 fighter jet with gray color').\nFor every instance, it has 2 F-16 fighter jets to its behind in view 3.\nIn view 1, how many 'F-16 fighter jet' that lie in the direct-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000908.png", "images/img_000910.png", "images/img_000903.png"], "supporting": {"draw_boxes": {"images/img_000903.png": []}}} +{"q_idx": "count_000468", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('woman in gray clothing').\nFor every instance, it has 1 woman in pink clothing to its behind in view 2.\nIn view 3, how many 'woman in gray clothing' that lie in the direct-left direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_001072.png", "images/img_001063.png", "images/img_001069.png"], "supporting": {"draw_boxes": {"images/img_001063.png": []}}} +{"q_idx": "count_000469", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('three balls snowman').\nFor every instance, it has 1 three balls snowman to its front, and 1 peach supercar to its left-front in view 2.\nIn view 1, how many 'peach supercar' that lie in the left-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000762.png", "images/img_000754.png"], "supporting": {"draw_boxes": {"images/img_000754.png": []}}} +{"q_idx": "count_000470", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('F-16 fighter jet').\nFor every instance, it has 2 airplane with white and orange liveries to its right-front in view 1.\nIn view 2, how many 'airplane with white and orange livery' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000282.png", "images/img_000290.png"], "supporting": {"draw_boxes": {"images/img_000282.png": []}}} +{"q_idx": "count_000471", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('man wearing green hood').\nFor every instance, it has 2 man wearing gray windbreakers to its right-front in view 2.\nIn view 1, how many 'man wearing gray windbreaker' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000027.png", "images/img_000019.png", "images/img_000028.png"], "supporting": {"draw_boxes": {"images/img_000019.png": []}}} +{"q_idx": "count_000472", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('man holding box').\nFor every instance, it has 1 black tombstone to its right in view 1.\nIn view 3, how many 'black tombstone' that lie in the direct-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000839.png", "images/img_000844.png", "images/img_000849.png"], "supporting": {"draw_boxes": {"images/img_000839.png": []}}} +{"q_idx": "count_000473", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('red painted metal bench with armrests').\nFor every instance, it has 1 basketball to its right-behind, and 1 basketball to its behind in view 2.\nIn view 1, how many 'basketball' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000589.png", "images/img_000581.png"], "supporting": {"draw_boxes": {"images/img_000581.png": []}}} +{"q_idx": "count_000474", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('bench made of black metal').\nFor every instance, it has 1 basketball to its right-behind in view 2.\nIn view 1, how many 'basketball' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000566.png", "images/img_000560.png", "images/img_000573.png"], "supporting": {"draw_boxes": {"images/img_000560.png": []}}} +{"q_idx": "count_000475", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('bench made of black metal').\nFor every instance, it has 1 soccer ball to its left-behind in view 1.\nIn view 2, how many 'soccer ball' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000628.png", "images/img_000638.png"], "supporting": {"draw_boxes": {"images/img_000628.png": []}}} +{"q_idx": "count_000476", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('green wooden bench with armrests').\nFor every instance, it has 2 baseballs to its left-front in view 2.\nIn view 3, how many 'baseball' that lie in the left-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000637.png", "images/img_000628.png", "images/img_000634.png"], "supporting": {"draw_boxes": {"images/img_000628.png": []}}} +{"q_idx": "count_000477", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('man wearing green hood').\nFor every instance, it has 1 man wearing gray windbreaker to its behind in view 2.\nIn view 3, how many 'man wearing gray windbreaker' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000024.png", "images/img_000020.png", "images/img_000027.png"], "supporting": {"draw_boxes": {"images/img_000020.png": []}}} +{"q_idx": "count_000478", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('stylized F22 fighter jet with white and red color').\nFor every instance, it has 1 airplane with red livery to its right-behind in view 2.\nIn view 1, how many 'airplane with red livery' that lie in the right-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000332.png", "images/img_000324.png"], "supporting": {"draw_boxes": {"images/img_000324.png": []}}} +{"q_idx": "count_000479", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('bench made of black metal').\nFor every instance, it has 1 basketball to its left-behind in view 3.\nIn view 2, how many 'basketball' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000639.png", "images/img_000636.png", "images/img_000628.png"], "supporting": {"draw_boxes": {"images/img_000628.png": []}}} +{"q_idx": "count_000480", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('natural wooden bench with armrests').\nFor every instance, it has 4 basketballs to its left-front, and 1 bench made of natural wood to its left in view 2.\nIn view 1, how many 'basketball' that lie in the left-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000605.png", "images/img_000598.png"], "supporting": {"draw_boxes": {"images/img_000598.png": []}}} +{"q_idx": "count_000481", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('lamp').\nFor every instance, it has 2 lamps to its right, and 1 beige stone steps to its left-behind in view 3.\nIn view 1, how many 'lamp' that lie in the direct-right direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000472.png", "images/img_000479.png", "images/img_000466.png"], "supporting": {"draw_boxes": {"images/img_000466.png": []}}} +{"q_idx": "count_000482", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('man wearing gray hat').\nFor every instance, it has 1 man wearing black coat to its front in view 1.\nIn view 3, how many 'man wearing black coat' that lie in the right-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000046.png", "images/img_000056.png", "images/img_000055.png"], "supporting": {"draw_boxes": {"images/img_000046.png": []}}} +{"q_idx": "count_000483", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('man wearing santa costume').\nFor every instance, it has 2 three balls snowmans to its left-behind in view 3.\nIn view 1, how many 'three balls snowman' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000763.png", "images/img_000769.png", "images/img_000755.png"], "supporting": {"draw_boxes": {"images/img_000755.png": []}}} +{"q_idx": "count_000484", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('flat tombstone').\nFor every instance, it has 2 man wearing a dark kurta and white pantses to its right in view 1.\nIn view 3, how many 'man wearing a dark kurta and white pants' that lie in the direct-right direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000839.png", "images/img_000845.png", "images/img_000848.png"], "supporting": {"draw_boxes": {"images/img_000839.png": []}}} +{"q_idx": "count_000485", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('bench made of natural wood').\nFor every instance, it has 5 beach balls to its right-front in view 2.\nIn view 1, how many 'beach ball' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000604.png", "images/img_000594.png"], "supporting": {"draw_boxes": {"images/img_000594.png": []}}} +{"q_idx": "count_000486", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('natural wooden bench with armrests').\nFor every instance, it has 5 beach balls to its right-front in view 2.\nIn view 1, how many 'beach ball' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000604.png", "images/img_000594.png"], "supporting": {"draw_boxes": {"images/img_000594.png": []}}} +{"q_idx": "count_000487", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('blue painted metal bench without armrests').\nFor every instance, it has 2 basketballs to its right-front, and 1 bench made of blue painted metal to its right in view 1.\nIn view 2, how many 'basketball' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000611.png", "images/img_000619.png"], "supporting": {"draw_boxes": {"images/img_000611.png": []}}} +{"q_idx": "count_000488", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('beige tombstone').\nFor every instance, it has 1 black upright tombstone with cross to its left in view 1.\nIn view 3, how many 'black upright tombstone with cross' that lie in the direct-left direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_001145.png", "images/img_001152.png", "images/img_001153.png"], "supporting": {"draw_boxes": {"images/img_001145.png": []}}} +{"q_idx": "count_000489", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('man holding box').\nFor every instance, it has 1 man wearing a green hoodie and orange shoes to its left in view 1.\nIn view 3, how many 'man wearing a green hoodie and orange shoes' that lie in the left-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000853.png", "images/img_000863.png", "images/img_000859.png"], "supporting": {"draw_boxes": {"images/img_000853.png": []}}} +{"q_idx": "count_000490", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('stylized F22 fighter jet with white and red color').\nFor every instance, it has 1 F-16 fighter jet to its left in view 1.\nIn view 2, how many 'stylized F22 fighter jet with white and red color' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000905.png", "images/img_000911.png"], "supporting": {"draw_boxes": {"images/img_000905.png": []}}} +{"q_idx": "count_000491", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('red painted metal bench with armrests').\nFor every instance, it has 1 basketball to its right-behind, and 1 basketball to its behind in view 1.\nIn view 2, how many 'basketball' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000581.png", "images/img_000587.png"], "supporting": {"draw_boxes": {"images/img_000581.png": []}}} +{"q_idx": "count_000492", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('red painted metal bench with armrests').\nFor every instance, it has 1 basketball to its right-behind, and 1 basketball to its behind in view 3.\nIn view 1, how many 'basketball' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000587.png", "images/img_000588.png", "images/img_000581.png"], "supporting": {"draw_boxes": {"images/img_000581.png": []}}} +{"q_idx": "count_000493", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('man in crouching posture').\nFor every instance, it has 1 tombstone without cross to its right in view 3.\nIn view 1, how many 'man in crouching posture' that lie in the direct-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000828.png", "images/img_000827.png", "images/img_000821.png"], "supporting": {"draw_boxes": {"images/img_000821.png": []}}} +{"q_idx": "count_000494", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('natural wooden bench with armrests').\nFor every instance, it has 5 basketballs to its left-behind, and 1 basketball to its right-behind in view 2.\nIn view 1, how many 'basketball' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000572.png", "images/img_000564.png"], "supporting": {"draw_boxes": {"images/img_000564.png": []}}} +{"q_idx": "count_000495", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('woman in gray clothing').\nFor every instance, it has 1 woman in gray clothing to its front, and 1 food stall with red and white striped roof to its left-front in view 3.\nIn view 2, how many 'food stall with red and white striped roof' that lie in the left-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_001073.png", "images/img_001072.png", "images/img_001066.png"], "supporting": {"draw_boxes": {"images/img_001066.png": []}}} +{"q_idx": "count_000496", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('bench made of natural wood').\nFor every instance, it has 5 basketballs to its left-behind, and 1 basketball to its right-behind in view 1.\nIn view 3, how many 'basketball' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000564.png", "images/img_000566.png", "images/img_000572.png"], "supporting": {"draw_boxes": {"images/img_000564.png": []}}} +{"q_idx": "count_000497", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('bench made of black metal').\nFor every instance, it has 1 basketball to its right-behind in view 1.\nIn view 2, how many 'basketball' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000560.png", "images/img_000570.png"], "supporting": {"draw_boxes": {"images/img_000560.png": []}}} +{"q_idx": "count_000498", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('bench made of natural wood').\nFor every instance, it has 5 beach balls to its right-front in view 1.\nIn view 2, how many 'beach ball' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000594.png", "images/img_000604.png", "images/img_000601.png"], "supporting": {"draw_boxes": {"images/img_000594.png": []}}} +{"q_idx": "count_000499", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('bench made of black metal').\nFor every instance, it has 1 basketball to its left-behind in view 3.\nIn view 1, how many 'basketball' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000638.png", "images/img_000639.png", "images/img_000628.png"], "supporting": {"draw_boxes": {"images/img_000628.png": []}}} +{"q_idx": "count_000500", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('F-22 fighter jet with gray color').\nFor every instance, it has 2 F-16 fighter jets to its behind in view 2.\nIn view 1, how many 'F-16 fighter jet' that lie in the direct-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000909.png", "images/img_000903.png"], "supporting": {"draw_boxes": {"images/img_000903.png": []}}} +{"q_idx": "count_000501", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('bench made of black metal').\nFor every instance, it has 2 basketballs to its right-front, 4 basketballs to its left-front, and 1 bench made of black metal to its left in view 1.\nIn view 3, how many 'basketball' that lie in the left-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000564.png", "images/img_000566.png", "images/img_000570.png"], "supporting": {"draw_boxes": {"images/img_000564.png": []}}} +{"q_idx": "count_000502", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('man wearing black coat').\nFor every instance, it has 2 man wearing green hoods to its behind in view 2.\nIn view 1, how many 'man wearing green hood' that lie in the direct-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000054.png", "images/img_000048.png"], "supporting": {"draw_boxes": {"images/img_000048.png": []}}} +{"q_idx": "count_000503", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('concrete mixer').\nFor every instance, it has 1 concrete mixer and 1 lamp to its right-behind in view 1.\nIn view 2, how many 'concrete mixer' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000435.png", "images/img_000443.png", "images/img_000439.png"], "supporting": {"draw_boxes": {"images/img_000435.png": []}}} +{"q_idx": "count_000504", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('woman in pink clothing').\nFor every instance, it has 1 woman in white clothing to its right in view 2.\nIn view 1, how many 'woman in white clothing' that lie in the direct-right direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_001056.png", "images/img_001049.png"], "supporting": {"draw_boxes": {"images/img_001049.png": []}}} +{"q_idx": "count_000505", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('man wearing orange coat').\nFor every instance, it has 2 snowman with black hat and stick armses to its left-front in view 3.\nIn view 1, how many 'snowman with black hat and stick arms' that lie in the left-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000750.png", "images/img_000749.png", "images/img_000740.png"], "supporting": {"draw_boxes": {"images/img_000740.png": []}}} +{"q_idx": "count_000506", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('bench made of black metal').\nFor every instance, it has 2 basketballs to its right-front, 4 basketballs to its left-front, and 1 bench made of black metal to its left in view 2.\nIn view 1, how many 'basketball' that lie in the left-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000570.png", "images/img_000564.png"], "supporting": {"draw_boxes": {"images/img_000564.png": []}}} +{"q_idx": "count_000507", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('airplane with light blue livery').\nFor every instance, it has 1 airplane with yellow livery to its right-behind in view 2.\nIn view 3, how many 'airplane with yellow livery' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000258.png", "images/img_000251.png", "images/img_000257.png"], "supporting": {"draw_boxes": {"images/img_000251.png": []}}} +{"q_idx": "count_000508", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('tombstone with pointed top').\nFor every instance, it has 1 beige upright tombstone with cross to its right in view 1.\nIn view 2, how many 'tombstone with pointed top' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_001127.png", "images/img_001135.png"], "supporting": {"draw_boxes": {"images/img_001127.png": []}}} +{"q_idx": "count_000509", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('green camouflage helicopter').\nFor every instance, it has 1 white and green striped hot air balloon to its right-behind in view 2.\nIn view 1, how many 'white and green striped hot air balloon' that lie in the right-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_001027.png", "images/img_001018.png"], "supporting": {"draw_boxes": {"images/img_001018.png": []}}} +{"q_idx": "count_000510", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('tombstone with pointed top').\nFor every instance, it has 1 man squatting to its right-front in view 1.\nIn view 2, how many 'tombstone with pointed top' that lie in the left-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_001173.png", "images/img_001182.png", "images/img_001185.png"], "supporting": {"draw_boxes": {"images/img_001173.png": []}}} +{"q_idx": "count_000511", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('airplane with red livery').\nFor every instance, it has 2 airplane with light blue liveries to its right in view 2.\nIn view 1, how many 'airplane with light blue livery' that lie in the direct-right direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000200.png", "images/img_000194.png", "images/img_000197.png"], "supporting": {"draw_boxes": {"images/img_000194.png": []}}} +{"q_idx": "count_000512", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('woman in pink clothing').\nFor every instance, it has 1 vendor cart with pink and white checkered roof to its left-front in view 1.\nIn view 3, how many 'woman in pink clothing' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_001095.png", "images/img_001107.png", "images/img_001101.png"], "supporting": {"draw_boxes": {"images/img_001095.png": []}}} +{"q_idx": "count_000513", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('mirage fighter jet').\nFor every instance, it has 2 harrier jumpjet fighter jets to its left in view 2.\nIn view 1, how many 'harrier jumpjet fighter jet' that lie in the left-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000944.png", "images/img_000936.png", "images/img_000938.png"], "supporting": {"draw_boxes": {"images/img_000936.png": []}}} +{"q_idx": "count_000514", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('F-16 fighter jet').\nFor every instance, it has 1 mirage fighter jet to its behind in view 1.\nIn view 3, how many 'mirage fighter jet' that lie in the direct-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000931.png", "images/img_000945.png", "images/img_000940.png"], "supporting": {"draw_boxes": {"images/img_000931.png": []}}} +{"q_idx": "count_000515", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('stone bench without armrests').\nFor every instance, it has 2 basketballs to its left-behind in view 2.\nIn view 1, how many 'basketball' that lie in the left-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000589.png", "images/img_000581.png", "images/img_000585.png"], "supporting": {"draw_boxes": {"images/img_000581.png": []}}} +{"q_idx": "count_000516", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('tombstone without cross').\nFor every instance, it has 1 colonnade with rectangular top to its left-behind in view 1.\nIn view 3, how many 'colonnade with rectangular top' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000858.png", "images/img_000860.png", "images/img_000866.png"], "supporting": {"draw_boxes": {"images/img_000858.png": []}}} +{"q_idx": "count_000517", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('bench made of natural wood').\nFor every instance, it has 1 beach ball to its left-front, 1 beach ball to its front, and 1 natural wooden bench with armrests to its left in view 2.\nIn view 3, how many 'beach ball' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000603.png", "images/img_000598.png", "images/img_000605.png"], "supporting": {"draw_boxes": {"images/img_000598.png": []}}} +{"q_idx": "count_000518", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('bench made of natural wood').\nFor every instance, it has 5 basketballs to its left-behind, and 1 basketball to its right-behind in view 1.\nIn view 2, how many 'basketball' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000564.png", "images/img_000573.png", "images/img_000566.png"], "supporting": {"draw_boxes": {"images/img_000564.png": []}}} +{"q_idx": "count_000519", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('bench made of natural wood').\nFor every instance, it has 5 beach balls to its right-front in view 2.\nIn view 1, how many 'beach ball' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000600.png", "images/img_000594.png", "images/img_000601.png"], "supporting": {"draw_boxes": {"images/img_000594.png": []}}} +{"q_idx": "count_000520", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('bench made of natural wood').\nFor every instance, it has 5 beach balls to its right-front in view 3.\nIn view 1, how many 'beach ball' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000600.png", "images/img_000607.png", "images/img_000594.png"], "supporting": {"draw_boxes": {"images/img_000594.png": []}}} +{"q_idx": "count_000521", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('man wearing black coat').\nFor every instance, it has 1 man wearing orange coat to its right in view 3.\nIn view 1, how many 'man wearing black coat' that lie in the left-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000009.png", "images/img_000013.png", "images/img_000002.png"], "supporting": {"draw_boxes": {"images/img_000002.png": []}}} +{"q_idx": "count_000522", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('red buttons snowman').\nFor every instance, it has 1 man wearing brown coat to its right-behind in view 2.\nIn view 3, how many 'man wearing brown coat' that lie in the left-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000753.png", "images/img_000743.png", "images/img_000749.png"], "supporting": {"draw_boxes": {"images/img_000743.png": []}}} +{"q_idx": "count_000523", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('bench made of black metal').\nFor every instance, it has 1 soccer ball to its left-behind in view 1.\nIn view 2, how many 'soccer ball' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000628.png", "images/img_000637.png", "images/img_000638.png"], "supporting": {"draw_boxes": {"images/img_000628.png": []}}} +{"q_idx": "count_000524", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('bench made of black metal').\nFor every instance, it has 1 tennis ball to its right-behind in view 2.\nIn view 1, how many 'tennis ball' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000603.png", "images/img_000594.png"], "supporting": {"draw_boxes": {"images/img_000594.png": []}}} +{"q_idx": "count_000525", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('bench made of black metal').\nFor every instance, it has 2 basketballs to its right-front, 4 basketballs to its left-front, and 1 bench made of black metal to its left in view 1.\nIn view 2, how many 'basketball' that lie in the left-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000564.png", "images/img_000570.png"], "supporting": {"draw_boxes": {"images/img_000564.png": []}}} +{"q_idx": "count_000526", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('green camouflage helicopter').\nFor every instance, it has 1 red and white helicopter to its front in view 3.\nIn view 2, how many 'green camouflage helicopter' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_001036.png", "images/img_001038.png", "images/img_001029.png"], "supporting": {"draw_boxes": {"images/img_001029.png": []}}} +{"q_idx": "count_000527", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('red painted metal bench with armrests').\nFor every instance, it has 1 basketball to its right-behind, and 1 basketball to its behind in view 2.\nIn view 1, how many 'basketball' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000590.png", "images/img_000581.png"], "supporting": {"draw_boxes": {"images/img_000581.png": []}}} +{"q_idx": "count_000528", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('man wearing gray windbreaker').\nFor every instance, it has 1 man wearing green hood to its front in view 1.\nIn view 2, how many 'man wearing gray windbreaker' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000020.png", "images/img_000025.png", "images/img_000023.png"], "supporting": {"draw_boxes": {"images/img_000020.png": []}}} +{"q_idx": "count_000529", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('bench made of natural wood').\nFor every instance, it has 1 beach ball to its left-front, 1 beach ball to its front, and 1 bench made of natural wood to its left in view 3.\nIn view 2, how many 'beach ball' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000600.png", "images/img_000605.png", "images/img_000598.png"], "supporting": {"draw_boxes": {"images/img_000598.png": []}}} +{"q_idx": "count_000530", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('bench made of black metal').\nFor every instance, it has 1 basketball to its right-behind in view 2.\nIn view 1, how many 'basketball' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000570.png", "images/img_000560.png", "images/img_000572.png"], "supporting": {"draw_boxes": {"images/img_000560.png": []}}} +{"q_idx": "count_000531", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('bench made of natural wood').\nFor every instance, it has 4 baseballs to its right-front, and 1 bench made of natural wood to its right in view 2.\nIn view 1, how many 'baseball' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000607.png", "images/img_000598.png"], "supporting": {"draw_boxes": {"images/img_000598.png": []}}} +{"q_idx": "count_000532", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('tombstone with pointed top').\nFor every instance, it has 2 white upright tombstone with crosses to its front in view 1.\nIn view 2, how many 'white upright tombstone with cross' that lie in the direct-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_001130.png", "images/img_001137.png"], "supporting": {"draw_boxes": {"images/img_001130.png": []}}} +{"q_idx": "count_000533", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('man wearing gray hat').\nFor every instance, it has 1 man wearing red jacket to its behind in view 2.\nIn view 1, how many 'man wearing gray hat' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000040.png", "images/img_000033.png"], "supporting": {"draw_boxes": {"images/img_000033.png": []}}} +{"q_idx": "count_000534", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('bench made of black metal').\nFor every instance, it has 1 bench made of blue painted metal to its right, and 1 bench made of blue painted metal to its left in view 1.\nIn view 2, how many 'bench made of blue painted metal' that lie in the direct-left direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000611.png", "images/img_000617.png", "images/img_000621.png"], "supporting": {"draw_boxes": {"images/img_000611.png": []}}} +{"q_idx": "count_000535", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('F-16 fighter jet').\nFor every instance, it has 2 F-22 fighter jet with gray colors to its right in view 2.\nIn view 1, how many 'F-22 fighter jet with gray color' that lie in the direct-right direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000956.png", "images/img_000948.png"], "supporting": {"draw_boxes": {"images/img_000948.png": []}}} +{"q_idx": "count_000536", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('natural wooden bench with armrests').\nFor every instance, it has 2 soccer balls to its left-behind in view 1.\nIn view 2, how many 'soccer ball' that lie in the left-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000564.png", "images/img_000572.png", "images/img_000566.png"], "supporting": {"draw_boxes": {"images/img_000564.png": []}}} +{"q_idx": "count_000537", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('natural wooden bench with armrests').\nFor every instance, it has 5 beach balls to its right-front in view 2.\nIn view 1, how many 'beach ball' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000603.png", "images/img_000594.png"], "supporting": {"draw_boxes": {"images/img_000594.png": []}}} +{"q_idx": "count_000538", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('green wooden bench with armrests').\nFor every instance, it has 2 basketballs to its left-behind in view 3.\nIn view 1, how many 'basketball' that lie in the left-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000622.png", "images/img_000618.png", "images/img_000615.png"], "supporting": {"draw_boxes": {"images/img_000615.png": []}}} +{"q_idx": "count_000539", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('man wearing brown coat').\nFor every instance, it has 2 man wearing green hoods to its behind in view 1.\nIn view 2, how many 'man wearing green hood' that lie in the direct-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000744.png", "images/img_000750.png", "images/img_000746.png"], "supporting": {"draw_boxes": {"images/img_000744.png": []}}} +{"q_idx": "count_000540", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('woman wearing underwear').\nFor every instance, it has 1 woman wearing underwear to its front, and 1 vendor cart with pink and white checkered roof to its right-front in view 2.\nIn view 3, how many 'woman wearing underwear' that lie in the direct-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_001123.png", "images/img_001115.png", "images/img_001122.png"], "supporting": {"draw_boxes": {"images/img_001115.png": []}}} +{"q_idx": "count_000541", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('F-22 fighter jet with gray color').\nFor every instance, it has 1 stylized F22 fighter jet with camoflague color to its left in view 2.\nIn view 1, how many 'stylized F22 fighter jet with camoflague color' that lie in the direct-left direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000894.png", "images/img_000886.png", "images/img_000895.png"], "supporting": {"draw_boxes": {"images/img_000886.png": []}}} +{"q_idx": "count_000542", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('man wearing black coat').\nFor every instance, it has 2 man wearing gray windbreakers to its left-front in view 2.\nIn view 1, how many 'man wearing gray windbreaker' that lie in the left-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000026.png", "images/img_000019.png"], "supporting": {"draw_boxes": {"images/img_000019.png": []}}} +{"q_idx": "count_000543", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('bench made of natural wood').\nFor every instance, it has 4 baseballs to its left-front, and 1 bench made of natural wood to its left in view 3.\nIn view 1, how many 'baseball' that lie in the left-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000604.png", "images/img_000605.png", "images/img_000598.png"], "supporting": {"draw_boxes": {"images/img_000598.png": []}}} +{"q_idx": "count_000544", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('F-16 fighter jet').\nFor every instance, it has 1 harrier jumpjet fighter jet to its behind in view 1.\nIn view 2, how many 'harrier jumpjet fighter jet' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000930.png", "images/img_000939.png"], "supporting": {"draw_boxes": {"images/img_000930.png": []}}} +{"q_idx": "count_000545", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('bench made of natural wood').\nFor every instance, it has 5 basketballs to its left-behind, and 1 basketball to its right-behind in view 2.\nIn view 1, how many 'basketball' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000570.png", "images/img_000564.png", "images/img_000569.png"], "supporting": {"draw_boxes": {"images/img_000564.png": []}}} +{"q_idx": "count_000546", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('red curved metal bench without armrests').\nFor every instance, it has 2 basketballs to its right-front in view 2.\nIn view 1, how many 'basketball' that lie in the right-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000590.png", "images/img_000581.png", "images/img_000587.png"], "supporting": {"draw_boxes": {"images/img_000581.png": []}}} +{"q_idx": "count_000547", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('worker wearing red hat and white shirt').\nFor every instance, it has 3 lamps to its left in view 1.\nIn view 2, how many 'lamp' that lie in the direct-left direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000466.png", "images/img_000472.png", "images/img_000478.png"], "supporting": {"draw_boxes": {"images/img_000466.png": []}}} +{"q_idx": "count_000548", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('bench made of natural wood').\nFor every instance, it has 5 basketballs to its left-behind, and 1 basketball to its right-behind in view 3.\nIn view 1, how many 'basketball' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000573.png", "images/img_000566.png", "images/img_000564.png"], "supporting": {"draw_boxes": {"images/img_000564.png": []}}} +{"q_idx": "count_000549", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('F-16 fighter jet').\nFor every instance, it has 1 stylized F22 fighter jet with grey and cyan color to its behind in view 1.\nIn view 2, how many 'stylized F22 fighter jet with grey and cyan color' that lie in the right-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000950.png", "images/img_000959.png"], "supporting": {"draw_boxes": {"images/img_000950.png": []}}} +{"q_idx": "count_000550", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('bench made of stone').\nFor every instance, it has 2 beach balls to its left-behind in view 1.\nIn view 2, how many 'beach ball' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000615.png", "images/img_000622.png"], "supporting": {"draw_boxes": {"images/img_000615.png": []}}} +{"q_idx": "count_000551", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('airplane with red livery').\nFor every instance, it has 1 airplane with white and orange livery to its behind in view 2.\nIn view 1, how many 'airplane with white and orange livery' that lie in the direct-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000239.png", "images/img_000232.png", "images/img_000242.png"], "supporting": {"draw_boxes": {"images/img_000232.png": []}}} +{"q_idx": "count_000552", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('man wearing santa costume').\nFor every instance, it has 2 no hat snowmans to its left-behind in view 1.\nIn view 2, how many 'no hat snowman' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000755.png", "images/img_000765.png"], "supporting": {"draw_boxes": {"images/img_000755.png": []}}} +{"q_idx": "count_000553", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('airplane with white and green livery').\nFor every instance, it has 1 airplane with red livery to its right in view 2.\nIn view 1, how many 'airplane with white and green livery' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000218.png", "images/img_000210.png", "images/img_000217.png"], "supporting": {"draw_boxes": {"images/img_000210.png": []}}} +{"q_idx": "count_000554", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('bench made of black metal').\nFor every instance, it has 2 basketballs to its right-front, 4 basketballs to its left-front, and 1 bench made of black metal to its left in view 2.\nIn view 3, how many 'basketball' that lie in the left-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000566.png", "images/img_000564.png", "images/img_000570.png"], "supporting": {"draw_boxes": {"images/img_000564.png": []}}} +{"q_idx": "count_000555", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('worker wearing red hat and white shirt').\nFor every instance, it has 3 lamps to its left in view 3.\nIn view 2, how many 'lamp' that lie in the direct-left direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000476.png", "images/img_000472.png", "images/img_000466.png"], "supporting": {"draw_boxes": {"images/img_000466.png": []}}} +{"q_idx": "count_000556", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('natural wooden bench with armrests').\nFor every instance, it has 5 basketballs to its left-behind, and 1 basketball to its right-behind in view 2.\nIn view 1, how many 'basketball' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000571.png", "images/img_000564.png", "images/img_000568.png"], "supporting": {"draw_boxes": {"images/img_000564.png": []}}} +{"q_idx": "count_000557", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('bench made of blue painted metal').\nFor every instance, it has 1 tennis ball to its left-front, and 1 blue painted metal bench without armrests to its right in view 2.\nIn view 1, how many 'tennis ball' that lie in the right-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000620.png", "images/img_000611.png"], "supporting": {"draw_boxes": {"images/img_000611.png": []}}} +{"q_idx": "count_000558", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('bench made of natural wood').\nFor every instance, it has 4 baseballs to its right-front, and 1 bench made of natural wood to its right in view 1.\nIn view 2, how many 'baseball' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000598.png", "images/img_000605.png"], "supporting": {"draw_boxes": {"images/img_000598.png": []}}} +{"q_idx": "count_000559", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('person with black hat').\nFor every instance, it has 1 person saluting to its left in view 1.\nIn view 2, how many 'person saluting' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000679.png", "images/img_000687.png"], "supporting": {"draw_boxes": {"images/img_000679.png": []}}} +{"q_idx": "count_000560", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('natural wooden bench with armrests').\nFor every instance, it has 1 beach ball to its left-front, 1 beach ball to its front, and 1 natural wooden bench with armrests to its left in view 2.\nIn view 3, how many 'beach ball' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000601.png", "images/img_000598.png", "images/img_000607.png"], "supporting": {"draw_boxes": {"images/img_000598.png": []}}} +{"q_idx": "count_000561", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('airplane with white and orange livery').\nFor every instance, it has 1 airplane with yellow livery to its left, and 1 airplane with yellow livery to its right in view 1.\nIn view 3, how many 'airplane with yellow livery' that lie in the direct-left direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000249.png", "images/img_000257.png", "images/img_000254.png"], "supporting": {"draw_boxes": {"images/img_000249.png": []}}} +{"q_idx": "count_000562", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('bench made of natural wood').\nFor every instance, it has 5 beach balls to its right-front in view 2.\nIn view 1, how many 'beach ball' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000600.png", "images/img_000594.png", "images/img_000604.png"], "supporting": {"draw_boxes": {"images/img_000594.png": []}}} +{"q_idx": "count_000563", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('airplane with white and green livery').\nFor every instance, it has 2 airplane with white and green liveries to its right in view 3.\nIn view 1, how many 'airplane with white and green livery' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000231.png", "images/img_000224.png", "images/img_000223.png"], "supporting": {"draw_boxes": {"images/img_000223.png": []}}} +{"q_idx": "count_000564", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('red curved metal bench without armrests').\nFor every instance, it has 2 basketballs to its right-front in view 2.\nIn view 3, how many 'basketball' that lie in the right-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000583.png", "images/img_000581.png", "images/img_000587.png"], "supporting": {"draw_boxes": {"images/img_000581.png": []}}} +{"q_idx": "count_000565", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('natural wooden bench with armrests').\nFor every instance, it has 5 beach balls to its right-front in view 2.\nIn view 3, how many 'beach ball' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000603.png", "images/img_000594.png", "images/img_000604.png"], "supporting": {"draw_boxes": {"images/img_000594.png": []}}} +{"q_idx": "count_000566", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('bench made of natural wood').\nFor every instance, it has 4 basketballs to its left-front, and 1 bench made of natural wood to its left in view 2.\nIn view 1, how many 'basketball' that lie in the left-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000607.png", "images/img_000598.png", "images/img_000603.png"], "supporting": {"draw_boxes": {"images/img_000598.png": []}}} +{"q_idx": "count_000567", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('man wearing green hood').\nFor every instance, it has 2 man wearing gray windbreakers to its right-front in view 1.\nIn view 3, how many 'man wearing gray windbreaker' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000019.png", "images/img_000022.png", "images/img_000028.png"], "supporting": {"draw_boxes": {"images/img_000019.png": []}}} +{"q_idx": "count_000568", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('bench made of black metal').\nFor every instance, it has 1 stone bench without armrests to its left in view 2.\nIn view 1, how many 'stone bench without armrests' that lie in the direct-left direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000587.png", "images/img_000577.png", "images/img_000586.png"], "supporting": {"draw_boxes": {"images/img_000577.png": []}}} +{"q_idx": "count_000569", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('natural wooden bench with armrests').\nFor every instance, it has 1 beach ball to its left-front, 1 beach ball to its front, and 1 natural wooden bench with armrests to its left in view 1.\nIn view 2, how many 'beach ball' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000598.png", "images/img_000605.png"], "supporting": {"draw_boxes": {"images/img_000598.png": []}}} +{"q_idx": "count_000570", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('natural wooden bench with armrests').\nFor every instance, it has 5 basketballs to its left-behind, and 1 basketball to its right-behind in view 1.\nIn view 2, how many 'basketball' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000564.png", "images/img_000570.png"], "supporting": {"draw_boxes": {"images/img_000564.png": []}}} +{"q_idx": "count_000571", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('airplane with red livery').\nFor every instance, it has 1 airplane with yellow livery to its front in view 1.\nIn view 2, how many 'airplane with red livery' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000268.png", "images/img_000274.png"], "supporting": {"draw_boxes": {"images/img_000268.png": []}}} +{"q_idx": "count_000572", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('natural wooden bench with armrests').\nFor every instance, it has 5 beach balls to its right-front in view 1.\nIn view 2, how many 'beach ball' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000594.png", "images/img_000600.png"], "supporting": {"draw_boxes": {"images/img_000594.png": []}}} +{"q_idx": "count_000573", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('man wearing red jacket').\nFor every instance, it has 1 striped scarf snowman to its right-behind in view 2.\nIn view 1, how many 'striped scarf snowman' that lie in the right-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000732.png", "images/img_000723.png", "images/img_000731.png"], "supporting": {"draw_boxes": {"images/img_000723.png": []}}} +{"q_idx": "count_000574", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('natural wooden bench with armrests').\nFor every instance, it has 2 basketballs to its right-front, and 1 basketball to its front in view 2.\nIn view 1, how many 'basketball' that lie in the right-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000602.png", "images/img_000594.png"], "supporting": {"draw_boxes": {"images/img_000594.png": []}}} +{"q_idx": "count_000575", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('surfboard').\nFor every instance, it has 1 beach ball to its left-front in view 2.\nIn view 1, how many 'surfboard' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000670.png", "images/img_000660.png"], "supporting": {"draw_boxes": {"images/img_000660.png": []}}} +{"q_idx": "count_000576", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('surfboard').\nFor every instance, it has 1 surfboard and 1 beach ball to its right in view 2.\nIn view 3, how many 'surfboard' that lie in the direct-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000670.png", "images/img_000663.png", "images/img_000671.png"], "supporting": {"draw_boxes": {"images/img_000663.png": []}}} +{"q_idx": "count_000577", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('F-16 fighter jet').\nFor every instance, it has 2 F-22 fighter jet with gray colors to its behind in view 1.\nIn view 2, how many 'F-22 fighter jet with gray color' that lie in the direct-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000311.png", "images/img_000319.png"], "supporting": {"draw_boxes": {"images/img_000311.png": []}}} +{"q_idx": "count_000578", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('airplane with white and orange livery').\nFor every instance, it has 1 airplane with white and green livery to its behind in view 3.\nIn view 2, how many 'airplane with white and green livery' that lie in the right-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000226.png", "images/img_000228.png", "images/img_000221.png"], "supporting": {"draw_boxes": {"images/img_000221.png": []}}} +{"q_idx": "count_000579", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('bench made of black metal').\nFor every instance, it has 2 basketballs to its right-front, 4 basketballs to its left-front, and 1 bench made of black metal to its left in view 3.\nIn view 1, how many 'basketball' that lie in the left-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000570.png", "images/img_000568.png", "images/img_000564.png"], "supporting": {"draw_boxes": {"images/img_000564.png": []}}} +{"q_idx": "count_000580", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('airplane with yellow livery').\nFor every instance, it has 1 airplane with light blue livery to its behind in view 1.\nIn view 2, how many 'airplane with light blue livery' that lie in the right-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 1, "image_paths": ["images/img_000311.png", "images/img_000317.png", "images/img_000314.png"], "supporting": {"draw_boxes": {"images/img_000311.png": []}}} +{"q_idx": "count_000581", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('stylized F22 fighter jet with grey and cyan color').\nFor every instance, it has 1 airplane with red livery to its right-behind in view 1.\nIn view 3, how many 'airplane with red livery' that lie in the right-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000267.png", "images/img_000271.png", "images/img_000276.png"], "supporting": {"draw_boxes": {"images/img_000267.png": []}}} +{"q_idx": "count_000582", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('airplane with yellow livery').\nFor every instance, it has 1 airplane with yellow livery and 1 airplane with white and green livery to its behind in view 2.\nIn view 3, how many 'airplane with white and green livery' that lie in the right-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000225.png", "images/img_000221.png", "images/img_000227.png"], "supporting": {"draw_boxes": {"images/img_000221.png": []}}} +{"q_idx": "count_000583", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('bench made of black metal').\nFor every instance, it has 1 soccer ball to its left-behind in view 1.\nIn view 2, how many 'soccer ball' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000628.png", "images/img_000636.png"], "supporting": {"draw_boxes": {"images/img_000628.png": []}}} +{"q_idx": "count_000584", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('airplane with white and green livery').\nFor every instance, it has 2 airplane with yellow liveries to its right in view 1.\nIn view 2, how many 'airplane with yellow livery' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000252.png", "images/img_000261.png"], "supporting": {"draw_boxes": {"images/img_000252.png": []}}} +{"q_idx": "count_000585", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('man wearing gray windbreaker').\nFor every instance, it has 1 man wearing black coat to its front in view 3.\nIn view 1, how many 'man wearing gray windbreaker' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000028.png", "images/img_000023.png", "images/img_000020.png"], "supporting": {"draw_boxes": {"images/img_000020.png": []}}} +{"q_idx": "count_000586", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('bench made of natural wood').\nFor every instance, it has 5 basketballs to its left-behind, and 1 basketball to its right-behind in view 1.\nIn view 2, how many 'basketball' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 5, "image_paths": ["images/img_000564.png", "images/img_000571.png"], "supporting": {"draw_boxes": {"images/img_000564.png": []}}} +{"q_idx": "count_000587", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('airplane with red livery').\nFor every instance, it has 1 airplane with yellow livery to its right, and 1 airplane with yellow livery to its left in view 3.\nIn view 2, how many 'airplane with yellow livery' that lie in the right-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_000259.png", "images/img_000260.png", "images/img_000252.png"], "supporting": {"draw_boxes": {"images/img_000252.png": []}}} +{"q_idx": "count_000588", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('white upright tombstone with cross').\nFor every instance, it has 2 black upright tombstone with crosses to its behind in view 2.\nIn view 1, how many 'black upright tombstone with cross' that lie in the direct-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_001167.png", "images/img_001160.png"], "supporting": {"draw_boxes": {"images/img_001160.png": []}}} +{"q_idx": "count_000589", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('woman in white clothing').\nFor every instance, it has 1 woman in pink clothing to its front in view 2.\nIn view 1, how many 'woman in white clothing' that lie in the left-behind direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 2, "image_paths": ["images/img_001056.png", "images/img_001048.png"], "supporting": {"draw_boxes": {"images/img_001048.png": []}}} +{"q_idx": "count_000590", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('natural wooden bench with armrests').\nFor every instance, it has 4 basketballs to its left-front, and 1 natural wooden bench with armrests to its left in view 1.\nIn view 2, how many 'basketball' that lie in the left-front direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 4, "image_paths": ["images/img_000598.png", "images/img_000607.png"], "supporting": {"draw_boxes": {"images/img_000598.png": []}}} +{"q_idx": "count_000591", "q_type": "count", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('worker wearing red hat and white shirt').\nFor every instance, it has 3 lamps to its left in view 1.\nIn view 3, how many 'lamp' that lie in the direct-left direction relative to these reference instances?", "options": {}, "question_prompt": "", "answer": 3, "image_paths": ["images/img_000466.png", "images/img_000474.png", "images/img_000472.png"], "supporting": {"draw_boxes": {"images/img_000466.png": []}}} diff --git a/detect/overall.jsonl b/detect/overall.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..bd8fff01496c33f1d7e9623688f4ac18624a5b71 --- /dev/null +++ b/detect/overall.jsonl @@ -0,0 +1,1600 @@ +{"q_idx": "detect_000001", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'stone bench without armrests'. For each such occluded instance of 'ball', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[731.0, 360.0, 743.0, 375.0], [725.0, 342.0, 743.0, 365.0]], "image_paths": ["images/img_000578.png", "images/img_000588.png", "images/img_000587.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000002", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'baseball' that are fully occluded by 'stone bench without armrests'. For each such occluded instance of 'baseball', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[636.0, 337.0, 646.0, 350.0]], "image_paths": ["images/img_000623.png", "images/img_000616.png", "images/img_000624.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000003", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'red shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[438.0, 297.0, 499.0, 345.0]], "image_paths": ["images/img_000524.png", "images/img_000525.png", "images/img_000518.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000004", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench made of red painted metal'. For each such occluded instance of 'ball', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[839.0, 442.0, 856.0, 463.0]], "image_paths": ["images/img_000636.png", "images/img_000637.png", "images/img_000628.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000005", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench made of blue painted metal'. For each such occluded instance of 'ball', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[815.0, 612.0, 839.0, 643.0], [795.0, 550.0, 807.0, 567.0], [806.0, 583.0, 821.0, 602.0], [752.0, 422.0, 771.0, 447.0], [810.0, 597.0, 830.0, 622.0], [761.0, 448.0, 781.0, 473.0], [775.0, 490.0, 796.0, 517.0]], "image_paths": ["images/img_000577.png", "images/img_000583.png", "images/img_000587.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000006", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[807.0, 457.0, 910.0, 535.0]], "image_paths": ["images/img_000484.png", "images/img_000494.png", "images/img_000492.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000007", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[284.0, 302.0, 324.0, 360.0]], "image_paths": ["images/img_000492.png", "images/img_000491.png", "images/img_000484.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000008", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'concrete mixer' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'concrete mixer', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[456.0, 237.0, 491.0, 293.0], [430.0, 252.0, 468.0, 305.0]], "image_paths": ["images/img_000439.png", "images/img_000431.png", "images/img_000441.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000009", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'baseball' that are fully occluded by 'bench made of stone'. For each such occluded instance of 'baseball', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[300.0, 370.0, 310.0, 383.0]], "image_paths": ["images/img_000615.png", "images/img_000621.png", "images/img_000622.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000010", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[111.0, 662.0, 239.0, 738.0]], "image_paths": ["images/img_000489.png", "images/img_000483.png", "images/img_000493.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000011", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stone steps'. For each such occluded instance of 'truck', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[155.0, 88.0, 370.0, 360.0]], "image_paths": ["images/img_000469.png", "images/img_000476.png", "images/img_000477.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000012", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'worker' that are fully occluded by 'stack of sandbags'. For each such occluded instance of 'worker', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[325.0, 287.0, 344.0, 355.0]], "image_paths": ["images/img_000409.png", "images/img_000401.png", "images/img_000411.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000013", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'white shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[355.0, 333.0, 414.0, 378.0]], "image_paths": ["images/img_000520.png", "images/img_000522.png", "images/img_000515.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000014", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[347.0, 367.0, 381.0, 435.0]], "image_paths": ["images/img_000553.png", "images/img_000552.png", "images/img_000546.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000015", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench made of white painted wood'. For each such occluded instance of 'ball', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[792.0, 513.0, 810.0, 535.0]], "image_paths": ["images/img_000595.png", "images/img_000601.png", "images/img_000604.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000016", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stone steps'. For each such occluded instance of 'truck', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[0.0, 405.0, 249.0, 880.0]], "image_paths": ["images/img_000475.png", "images/img_000479.png", "images/img_000469.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000017", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'natural wooden bench with armrests'. For each such occluded instance of 'ball', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[133.0, 447.0, 149.0, 467.0]], "image_paths": ["images/img_000571.png", "images/img_000565.png", "images/img_000572.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000018", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'worker' that are fully occluded by 'stack of sandbags'. For each such occluded instance of 'worker', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[354.0, 290.0, 375.0, 350.0]], "image_paths": ["images/img_000433.png", "images/img_000439.png", "images/img_000442.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000019", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'forklift' that are fully occluded by 'green shipping container'. For each such occluded instance of 'forklift', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[527.0, 320.0, 542.0, 365.0]], "image_paths": ["images/img_000522.png", "images/img_000515.png", "images/img_000520.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000020", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'black stack of bricks'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[819.0, 383.0, 1000.0, 500.0]], "image_paths": ["images/img_000458.png", "images/img_000459.png", "images/img_000450.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000021", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[119.0, 437.0, 194.0, 513.0]], "image_paths": ["images/img_000484.png", "images/img_000492.png", "images/img_000490.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000022", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'baseball' that are fully occluded by 'stone bench without armrests'. For each such occluded instance of 'baseball', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[263.0, 373.0, 274.0, 388.0]], "image_paths": ["images/img_000581.png", "images/img_000588.png", "images/img_000590.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000023", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'blue painted metal bench without armrests'. For each such occluded instance of 'ball', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[761.0, 448.0, 781.0, 473.0], [775.0, 490.0, 796.0, 517.0], [752.0, 422.0, 771.0, 447.0]], "image_paths": ["images/img_000578.png", "images/img_000588.png", "images/img_000587.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000024", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'concrete mixer' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'concrete mixer', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[247.0, 298.0, 269.0, 360.0], [237.0, 325.0, 266.0, 387.0]], "image_paths": ["images/img_000432.png", "images/img_000441.png", "images/img_000438.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000025", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'worker' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'worker', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[468.0, 270.0, 486.0, 332.0]], "image_paths": ["images/img_000415.png", "images/img_000414.png", "images/img_000407.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000026", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'forklift' that are fully occluded by 'green shipping container'. For each such occluded instance of 'forklift', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[366.0, 340.0, 403.0, 387.0]], "image_paths": ["images/img_000515.png", "images/img_000521.png", "images/img_000524.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000027", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'worker' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'worker', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[215.0, 397.0, 242.0, 470.0]], "image_paths": ["images/img_000448.png", "images/img_000455.png", "images/img_000457.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000028", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[119.0, 437.0, 194.0, 513.0]], "image_paths": ["images/img_000484.png", "images/img_000490.png", "images/img_000494.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000029", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'green shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[113.0, 427.0, 206.0, 548.0]], "image_paths": ["images/img_000541.png", "images/img_000534.png", "images/img_000543.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000030", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench'. For each such occluded instance of 'ball', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[710.0, 375.0, 724.0, 393.0], [835.0, 438.0, 853.0, 460.0]], "image_paths": ["images/img_000641.png", "images/img_000639.png", "images/img_000631.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000031", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench made of natural wood'. For each such occluded instance of 'ball', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[460.0, 280.0, 468.0, 292.0]], "image_paths": ["images/img_000597.png", "images/img_000605.png", "images/img_000603.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000032", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'white shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[556.0, 300.0, 609.0, 345.0]], "image_paths": ["images/img_000487.png", "images/img_000496.png", "images/img_000495.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000033", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[450.0, 287.0, 486.0, 383.0]], "image_paths": ["images/img_000543.png", "images/img_000542.png", "images/img_000533.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000034", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'black stack of bricks'. For each such occluded instance of 'truck', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[441.0, 190.0, 514.0, 293.0]], "image_paths": ["images/img_000450.png", "images/img_000460.png", "images/img_000457.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000035", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'white shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[625.0, 635.0, 711.0, 702.0]], "image_paths": ["images/img_000515.png", "images/img_000523.png", "images/img_000524.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000036", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[675.0, 312.0, 757.0, 360.0]], "image_paths": ["images/img_000493.png", "images/img_000484.png", "images/img_000491.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000037", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stone steps'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[155.0, 88.0, 370.0, 360.0]], "image_paths": ["images/img_000479.png", "images/img_000477.png", "images/img_000471.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000038", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'truck', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[655.0, 227.0, 789.0, 345.0]], "image_paths": ["images/img_000457.png", "images/img_000448.png", "images/img_000458.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000039", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'concrete mixer' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'concrete mixer', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[676.0, 282.0, 718.0, 335.0], [640.0, 282.0, 679.0, 335.0]], "image_paths": ["images/img_000439.png", "images/img_000440.png", "images/img_000431.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000040", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'natural wooden bench with armrests'. For each such occluded instance of 'ball', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[105.0, 517.0, 117.0, 532.0]], "image_paths": ["images/img_000597.png", "images/img_000603.png", "images/img_000604.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000041", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[454.0, 328.0, 511.0, 393.0]], "image_paths": ["images/img_000553.png", "images/img_000551.png", "images/img_000546.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000042", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'white shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[264.0, 352.0, 340.0, 397.0]], "image_paths": ["images/img_000524.png", "images/img_000517.png", "images/img_000522.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000043", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench made of natural wood'. For each such occluded instance of 'ball', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[226.0, 348.0, 236.0, 360.0]], "image_paths": ["images/img_000604.png", "images/img_000607.png", "images/img_000597.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000044", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'white shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[158.0, 567.0, 242.0, 617.0]], "image_paths": ["images/img_000487.png", "images/img_000493.png", "images/img_000496.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000045", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[480.0, 260.0, 546.0, 312.0]], "image_paths": ["images/img_000483.png", "images/img_000492.png", "images/img_000493.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000046", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'black stack of bricks'. For each such occluded instance of 'truck', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[819.0, 383.0, 1000.0, 500.0]], "image_paths": ["images/img_000459.png", "images/img_000450.png", "images/img_000457.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000047", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'green shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[450.0, 240.0, 526.0, 325.0]], "image_paths": ["images/img_000534.png", "images/img_000543.png", "images/img_000541.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000048", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'worker' that are fully occluded by 'stack of sandbags'. For each such occluded instance of 'worker', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[795.0, 467.0, 824.0, 550.0]], "image_paths": ["images/img_000442.png", "images/img_000432.png", "images/img_000438.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000049", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stack of sandbags'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[698.0, 242.0, 833.0, 365.0]], "image_paths": ["images/img_000433.png", "images/img_000441.png", "images/img_000440.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000050", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'green shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[450.0, 240.0, 526.0, 325.0]], "image_paths": ["images/img_000543.png", "images/img_000542.png", "images/img_000534.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000051", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'forklift' that are fully occluded by 'red shipping container'. For each such occluded instance of 'forklift', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[765.0, 493.0, 821.0, 552.0]], "image_paths": ["images/img_000522.png", "images/img_000525.png", "images/img_000517.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000052", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'stone bench without armrests'. For each such occluded instance of 'ball', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[188.0, 467.0, 207.0, 493.0], [185.0, 485.0, 203.0, 507.0]], "image_paths": ["images/img_000587.png", "images/img_000586.png", "images/img_000580.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000053", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[206.0, 450.0, 321.0, 542.0]], "image_paths": ["images/img_000533.png", "images/img_000540.png", "images/img_000542.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000054", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stack of sandbags'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[491.0, 192.0, 564.0, 297.0]], "image_paths": ["images/img_000432.png", "images/img_000440.png", "images/img_000439.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000055", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'white shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[620.0, 368.0, 651.0, 433.0]], "image_paths": ["images/img_000554.png", "images/img_000549.png", "images/img_000556.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000056", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'forklift' that are fully occluded by 'green shipping container'. For each such occluded instance of 'forklift', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[656.0, 373.0, 693.0, 423.0]], "image_paths": ["images/img_000523.png", "images/img_000515.png", "images/img_000521.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000057", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'trolley' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'trolley', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[176.0, 605.0, 255.0, 668.0]], "image_paths": ["images/img_000412.png", "images/img_000405.png", "images/img_000411.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000058", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench made of natural wood'. For each such occluded instance of 'ball', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[840.0, 450.0, 851.0, 463.0]], "image_paths": ["images/img_000597.png", "images/img_000607.png", "images/img_000604.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000059", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench'. For each such occluded instance of 'ball', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[710.0, 375.0, 724.0, 393.0], [835.0, 438.0, 853.0, 460.0]], "image_paths": ["images/img_000638.png", "images/img_000631.png", "images/img_000641.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000060", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench made of blue painted metal'. For each such occluded instance of 'ball', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[230.0, 370.0, 244.0, 388.0], [239.0, 355.0, 247.0, 367.0]], "image_paths": ["images/img_000611.png", "images/img_000617.png", "images/img_000621.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000061", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'white shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[750.0, 475.0, 836.0, 583.0], [481.0, 317.0, 556.0, 397.0]], "image_paths": ["images/img_000496.png", "images/img_000494.png", "images/img_000487.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000062", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench made of natural wood'. For each such occluded instance of 'ball', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[105.0, 517.0, 117.0, 532.0]], "image_paths": ["images/img_000597.png", "images/img_000607.png", "images/img_000603.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000063", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench made of natural wood'. For each such occluded instance of 'ball', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[460.0, 280.0, 468.0, 292.0]], "image_paths": ["images/img_000605.png", "images/img_000607.png", "images/img_000598.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000064", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'worker' that are fully occluded by 'stack of sandbags'. For each such occluded instance of 'worker', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[797.0, 450.0, 828.0, 522.0]], "image_paths": ["images/img_000412.png", "images/img_000403.png", "images/img_000410.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000065", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'stone bench without armrests'. For each such occluded instance of 'ball', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[311.0, 357.0, 326.0, 377.0], [300.0, 370.0, 310.0, 383.0], [339.0, 342.0, 354.0, 362.0], [326.0, 352.0, 340.0, 370.0]], "image_paths": ["images/img_000616.png", "images/img_000623.png", "images/img_000622.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000066", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench with armrests'. For each such occluded instance of 'ball', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[453.0, 303.0, 466.0, 322.0], [519.0, 310.0, 527.0, 322.0], [438.0, 303.0, 451.0, 322.0]], "image_paths": ["images/img_000570.png", "images/img_000561.png", "images/img_000568.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000067", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench with armrests'. For each such occluded instance of 'ball', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[411.0, 312.0, 426.0, 332.0]], "image_paths": ["images/img_000618.png", "images/img_000619.png", "images/img_000612.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000068", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stack of sandbags'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[698.0, 242.0, 833.0, 365.0]], "image_paths": ["images/img_000439.png", "images/img_000441.png", "images/img_000431.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000069", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'forklift' that are fully occluded by 'red shipping container'. For each such occluded instance of 'forklift', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[765.0, 493.0, 821.0, 552.0]], "image_paths": ["images/img_000517.png", "images/img_000525.png", "images/img_000522.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000070", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stack of sandbags'. For each such occluded instance of 'truck', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[698.0, 242.0, 833.0, 365.0]], "image_paths": ["images/img_000442.png", "images/img_000433.png", "images/img_000441.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000071", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stack of sandbags'. For each such occluded instance of 'truck', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[491.0, 192.0, 564.0, 297.0]], "image_paths": ["images/img_000440.png", "images/img_000439.png", "images/img_000433.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000072", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stack of sandbags'. For each such occluded instance of 'truck', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[491.0, 192.0, 564.0, 297.0]], "image_paths": ["images/img_000440.png", "images/img_000432.png", "images/img_000439.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000073", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'truck', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[458.0, 172.0, 532.0, 280.0]], "image_paths": ["images/img_000420.png", "images/img_000426.png", "images/img_000428.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000074", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'green shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[594.0, 343.0, 634.0, 427.0]], "image_paths": ["images/img_000483.png", "images/img_000491.png", "images/img_000489.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000075", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'red shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[576.0, 308.0, 657.0, 405.0]], "image_paths": ["images/img_000541.png", "images/img_000543.png", "images/img_000535.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000076", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[675.0, 312.0, 757.0, 360.0]], "image_paths": ["images/img_000493.png", "images/img_000494.png", "images/img_000484.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000077", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench made of red painted metal'. For each such occluded instance of 'ball', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[677.0, 292.0, 693.0, 312.0]], "image_paths": ["images/img_000637.png", "images/img_000627.png", "images/img_000636.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000078", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'baseball' that are fully occluded by 'bench made of blue painted metal'. For each such occluded instance of 'baseball', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[603.0, 302.0, 613.0, 315.0]], "image_paths": ["images/img_000576.png", "images/img_000583.png", "images/img_000585.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000079", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'red curved metal bench without armrests'. For each such occluded instance of 'ball', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[230.0, 340.0, 245.0, 360.0]], "image_paths": ["images/img_000634.png", "images/img_000628.png", "images/img_000635.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000080", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'worker' that are fully occluded by 'stack of sandbags'. For each such occluded instance of 'worker', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[354.0, 290.0, 375.0, 350.0]], "image_paths": ["images/img_000432.png", "images/img_000439.png", "images/img_000441.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000081", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'truck', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[0.0, 405.0, 145.0, 533.0]], "image_paths": ["images/img_000426.png", "images/img_000428.png", "images/img_000420.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000082", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'truck', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[458.0, 172.0, 532.0, 280.0]], "image_paths": ["images/img_000428.png", "images/img_000421.png", "images/img_000427.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000083", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'red curved metal bench without armrests'. For each such occluded instance of 'ball', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[839.0, 442.0, 856.0, 463.0]], "image_paths": ["images/img_000628.png", "images/img_000637.png", "images/img_000634.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000084", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'concrete mixer' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'concrete mixer', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[456.0, 237.0, 491.0, 293.0], [430.0, 252.0, 468.0, 305.0]], "image_paths": ["images/img_000432.png", "images/img_000438.png", "images/img_000439.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000085", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[347.0, 367.0, 381.0, 435.0]], "image_paths": ["images/img_000552.png", "images/img_000553.png", "images/img_000546.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000086", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stone steps'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[0.0, 405.0, 249.0, 880.0]], "image_paths": ["images/img_000477.png", "images/img_000475.png", "images/img_000469.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000087", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'baseball' that are fully occluded by 'green wooden bench with armrests'. For each such occluded instance of 'baseball', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[791.0, 523.0, 805.0, 540.0]], "image_paths": ["images/img_000560.png", "images/img_000570.png", "images/img_000568.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000088", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'white painted wooden bench with armrests'. For each such occluded instance of 'ball', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[525.0, 298.0, 537.0, 315.0]], "image_paths": ["images/img_000602.png", "images/img_000604.png", "images/img_000595.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000089", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'forklift' that are fully occluded by 'stack of sandbags'. For each such occluded instance of 'forklift', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[819.0, 362.0, 1000.0, 512.0]], "image_paths": ["images/img_000412.png", "images/img_000403.png", "images/img_000410.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000090", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'baseball' that are fully occluded by 'stone bench without armrests'. For each such occluded instance of 'baseball', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[624.0, 323.0, 634.0, 337.0]], "image_paths": ["images/img_000587.png", "images/img_000590.png", "images/img_000581.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000091", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench made of stone'. For each such occluded instance of 'ball', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[672.0, 350.0, 688.0, 370.0], [701.0, 372.0, 710.0, 383.0], [660.0, 345.0, 674.0, 362.0]], "image_paths": ["images/img_000614.png", "images/img_000624.png", "images/img_000620.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000092", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench made of natural wood'. For each such occluded instance of 'ball', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[460.0, 280.0, 468.0, 292.0]], "image_paths": ["images/img_000607.png", "images/img_000605.png", "images/img_000598.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000093", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[271.0, 460.0, 336.0, 542.0]], "image_paths": ["images/img_000546.png", "images/img_000550.png", "images/img_000551.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000094", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'green shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[113.0, 427.0, 206.0, 548.0]], "image_paths": ["images/img_000534.png", "images/img_000543.png", "images/img_000541.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000095", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[576.0, 308.0, 657.0, 405.0]], "image_paths": ["images/img_000542.png", "images/img_000543.png", "images/img_000533.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000096", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'trolley' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'trolley', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[180.0, 432.0, 249.0, 475.0]], "image_paths": ["images/img_000412.png", "images/img_000405.png", "images/img_000413.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000097", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'truck', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[672.0, 223.0, 810.0, 347.0]], "image_paths": ["images/img_000429.png", "images/img_000425.png", "images/img_000419.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000098", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'truck', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[458.0, 172.0, 532.0, 280.0]], "image_paths": ["images/img_000428.png", "images/img_000419.png", "images/img_000427.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000099", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'truck', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[662.0, 210.0, 838.0, 360.0]], "image_paths": ["images/img_000405.png", "images/img_000414.png", "images/img_000415.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000100", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'white shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[364.0, 363.0, 420.0, 407.0]], "image_paths": ["images/img_000549.png", "images/img_000555.png", "images/img_000554.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000101", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'worker' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'worker', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[767.0, 420.0, 804.0, 490.0]], "image_paths": ["images/img_000425.png", "images/img_000417.png", "images/img_000426.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000102", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[271.0, 460.0, 336.0, 542.0]], "image_paths": ["images/img_000553.png", "images/img_000546.png", "images/img_000551.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000103", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'worker' that are fully occluded by 'stack of sandbags'. For each such occluded instance of 'worker', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[185.0, 415.0, 215.0, 488.0]], "image_paths": ["images/img_000402.png", "images/img_000408.png", "images/img_000412.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000104", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'white shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[364.0, 363.0, 420.0, 407.0]], "image_paths": ["images/img_000549.png", "images/img_000554.png", "images/img_000557.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000105", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[292.0, 333.0, 376.0, 435.0]], "image_paths": ["images/img_000533.png", "images/img_000541.png", "images/img_000542.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000106", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench'. For each such occluded instance of 'ball', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[220.0, 437.0, 236.0, 457.0], [249.0, 342.0, 264.0, 362.0], [136.0, 688.0, 160.0, 718.0]], "image_paths": ["images/img_000632.png", "images/img_000638.png", "images/img_000639.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000107", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'worker' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'worker', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[468.0, 268.0, 487.0, 333.0]], "image_paths": ["images/img_000426.png", "images/img_000424.png", "images/img_000418.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000108", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'truck', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[458.0, 172.0, 532.0, 280.0]], "image_paths": ["images/img_000425.png", "images/img_000419.png", "images/img_000428.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000109", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stone steps'. For each such occluded instance of 'truck', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[155.0, 88.0, 370.0, 360.0]], "image_paths": ["images/img_000477.png", "images/img_000469.png", "images/img_000475.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000110", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench with armrests'. For each such occluded instance of 'ball', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[796.0, 410.0, 816.0, 435.0], [833.0, 432.0, 849.0, 453.0]], "image_paths": ["images/img_000590.png", "images/img_000589.png", "images/img_000580.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000111", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[206.0, 450.0, 321.0, 542.0]], "image_paths": ["images/img_000543.png", "images/img_000540.png", "images/img_000533.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000112", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench with armrests'. For each such occluded instance of 'ball', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[182.0, 485.0, 203.0, 510.0], [200.0, 430.0, 219.0, 455.0], [193.0, 480.0, 204.0, 495.0], [179.0, 502.0, 198.0, 527.0], [209.0, 407.0, 226.0, 428.0]], "image_paths": ["images/img_000634.png", "images/img_000627.png", "images/img_000635.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000113", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'forklift' that are fully occluded by 'red shipping container'. For each such occluded instance of 'forklift', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[676.0, 353.0, 725.0, 393.0]], "image_paths": ["images/img_000526.png", "images/img_000517.png", "images/img_000524.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000114", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'white shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[689.0, 338.0, 736.0, 423.0], [369.0, 338.0, 407.0, 423.0]], "image_paths": ["images/img_000495.png", "images/img_000487.png", "images/img_000494.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000115", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'forklift' that are fully occluded by 'green shipping container'. For each such occluded instance of 'forklift', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[366.0, 340.0, 403.0, 387.0]], "image_paths": ["images/img_000515.png", "images/img_000520.png", "images/img_000521.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000116", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'trolley' that are fully occluded by 'stack of sandbags'. For each such occluded instance of 'trolley', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[235.0, 350.0, 266.0, 388.0]], "image_paths": ["images/img_000455.png", "images/img_000449.png", "images/img_000458.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000117", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'black stack of bricks'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[0.0, 430.0, 164.0, 555.0]], "image_paths": ["images/img_000449.png", "images/img_000455.png", "images/img_000458.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000118", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'forklift' that are fully occluded by 'green shipping container'. For each such occluded instance of 'forklift', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[656.0, 373.0, 693.0, 423.0]], "image_paths": ["images/img_000515.png", "images/img_000523.png", "images/img_000520.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000119", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stone steps'. For each such occluded instance of 'truck', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[0.0, 185.0, 214.0, 510.0]], "image_paths": ["images/img_000475.png", "images/img_000469.png", "images/img_000476.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000120", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'concrete mixer' that are fully occluded by 'stack of sandbags'. For each such occluded instance of 'concrete mixer', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[255.0, 297.0, 276.0, 357.0]], "image_paths": ["images/img_000453.png", "images/img_000459.png", "images/img_000461.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000121", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'worker' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'worker', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[537.0, 275.0, 561.0, 332.0]], "image_paths": ["images/img_000457.png", "images/img_000455.png", "images/img_000448.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000122", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'white shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[620.0, 368.0, 651.0, 433.0]], "image_paths": ["images/img_000549.png", "images/img_000557.png", "images/img_000556.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000123", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stone steps'. For each such occluded instance of 'truck', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[155.0, 88.0, 370.0, 360.0]], "image_paths": ["images/img_000477.png", "images/img_000469.png", "images/img_000478.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000124", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'worker' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'worker', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[198.0, 447.0, 221.0, 525.0]], "image_paths": ["images/img_000415.png", "images/img_000405.png", "images/img_000412.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000125", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'natural wooden bench with armrests'. For each such occluded instance of 'ball', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[105.0, 517.0, 117.0, 532.0]], "image_paths": ["images/img_000603.png", "images/img_000605.png", "images/img_000597.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000126", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'basketball' that are fully occluded by 'blue painted metal bench without armrests'. For each such occluded instance of 'basketball', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[733.0, 352.0, 749.0, 373.0], [736.0, 362.0, 754.0, 383.0]], "image_paths": ["images/img_000570.png", "images/img_000561.png", "images/img_000571.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000127", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'natural wooden bench with armrests'. For each such occluded instance of 'ball', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[460.0, 280.0, 468.0, 292.0]], "image_paths": ["images/img_000605.png", "images/img_000604.png", "images/img_000598.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000128", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[318.0, 582.0, 399.0, 642.0]], "image_paths": ["images/img_000546.png", "images/img_000550.png", "images/img_000554.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000129", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'red shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[626.0, 322.0, 659.0, 375.0]], "image_paths": ["images/img_000526.png", "images/img_000518.png", "images/img_000525.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000130", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'truck', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[166.0, 250.0, 301.0, 373.0]], "image_paths": ["images/img_000458.png", "images/img_000448.png", "images/img_000456.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000131", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'natural wooden bench with armrests'. For each such occluded instance of 'ball', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[325.0, 312.0, 338.0, 328.0]], "image_paths": ["images/img_000571.png", "images/img_000572.png", "images/img_000565.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000132", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'natural wooden bench with armrests'. For each such occluded instance of 'ball', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[133.0, 447.0, 149.0, 467.0]], "image_paths": ["images/img_000571.png", "images/img_000564.png", "images/img_000572.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000133", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'green shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[594.0, 343.0, 634.0, 427.0]], "image_paths": ["images/img_000491.png", "images/img_000493.png", "images/img_000483.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000134", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'trolley' that are fully occluded by 'stone steps'. For each such occluded instance of 'trolley', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[110.0, 653.0, 164.0, 715.0]], "image_paths": ["images/img_000422.png", "images/img_000417.png", "images/img_000423.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000135", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[206.0, 450.0, 321.0, 542.0]], "image_paths": ["images/img_000543.png", "images/img_000533.png", "images/img_000540.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000136", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'basketball' that are fully occluded by 'green wooden bench with armrests'. For each such occluded instance of 'basketball', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[796.0, 468.0, 816.0, 493.0], [818.0, 530.0, 839.0, 557.0], [777.0, 417.0, 796.0, 440.0]], "image_paths": ["images/img_000636.png", "images/img_000638.png", "images/img_000628.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000137", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'trolley' that are fully occluded by 'stone steps'. For each such occluded instance of 'trolley', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[168.0, 415.0, 215.0, 462.0]], "image_paths": ["images/img_000477.png", "images/img_000476.png", "images/img_000470.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000138", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stone steps'. For each such occluded instance of 'truck', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[155.0, 88.0, 370.0, 360.0]], "image_paths": ["images/img_000477.png", "images/img_000478.png", "images/img_000469.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000139", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'worker' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'worker', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[296.0, 318.0, 315.0, 383.0]], "image_paths": ["images/img_000411.png", "images/img_000413.png", "images/img_000405.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000140", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'stone bench without armrests'. For each such occluded instance of 'ball', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[236.0, 648.0, 258.0, 675.0], [217.0, 630.0, 230.0, 647.0], [250.0, 665.0, 269.0, 690.0]], "image_paths": ["images/img_000620.png", "images/img_000614.png", "images/img_000623.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000141", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stone steps'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[471.0, 73.0, 593.0, 315.0]], "image_paths": ["images/img_000469.png", "images/img_000478.png", "images/img_000476.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000142", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'worker' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'worker', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[198.0, 447.0, 221.0, 525.0]], "image_paths": ["images/img_000412.png", "images/img_000414.png", "images/img_000406.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000143", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'red shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[432.0, 263.0, 510.0, 348.0], [292.0, 333.0, 376.0, 435.0]], "image_paths": ["images/img_000540.png", "images/img_000534.png", "images/img_000541.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000144", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'red shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[292.0, 333.0, 376.0, 435.0]], "image_paths": ["images/img_000535.png", "images/img_000541.png", "images/img_000542.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000145", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'baseball' that are fully occluded by 'green wooden bench with armrests'. For each such occluded instance of 'baseball', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[346.0, 332.0, 356.0, 345.0]], "image_paths": ["images/img_000559.png", "images/img_000566.png", "images/img_000567.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000146", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'red shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[626.0, 322.0, 659.0, 375.0]], "image_paths": ["images/img_000526.png", "images/img_000527.png", "images/img_000518.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000147", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'black stack of bricks'. For each such occluded instance of 'truck', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[799.0, 622.0, 1000.0, 863.0]], "image_paths": ["images/img_000460.png", "images/img_000458.png", "images/img_000450.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000148", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench with armrests'. For each such occluded instance of 'ball', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[796.0, 410.0, 816.0, 435.0], [833.0, 432.0, 849.0, 453.0]], "image_paths": ["images/img_000580.png", "images/img_000590.png", "images/img_000589.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000149", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'blue painted metal bench without armrests'. For each such occluded instance of 'ball', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[815.0, 612.0, 839.0, 643.0], [795.0, 550.0, 807.0, 567.0], [752.0, 422.0, 771.0, 447.0], [806.0, 583.0, 821.0, 602.0], [810.0, 597.0, 830.0, 622.0], [761.0, 448.0, 781.0, 473.0], [775.0, 490.0, 796.0, 517.0]], "image_paths": ["images/img_000583.png", "images/img_000577.png", "images/img_000587.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000150", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'concrete mixer' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'concrete mixer', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[456.0, 237.0, 491.0, 293.0], [430.0, 252.0, 468.0, 305.0]], "image_paths": ["images/img_000439.png", "images/img_000432.png", "images/img_000438.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000151", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'forklift' that are fully occluded by 'green shipping container'. For each such occluded instance of 'forklift', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[656.0, 373.0, 693.0, 423.0]], "image_paths": ["images/img_000515.png", "images/img_000522.png", "images/img_000523.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000152", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'white shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[776.0, 437.0, 891.0, 518.0]], "image_paths": ["images/img_000529.png", "images/img_000537.png", "images/img_000538.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000153", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'trolley' that are fully occluded by 'stone steps'. For each such occluded instance of 'trolley', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[733.0, 313.0, 762.0, 370.0]], "image_paths": ["images/img_000418.png", "images/img_000424.png", "images/img_000426.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000154", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[0.0, 405.0, 145.0, 533.0]], "image_paths": ["images/img_000429.png", "images/img_000426.png", "images/img_000419.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000155", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench made of black metal'. For each such occluded instance of 'ball', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[316.0, 295.0, 329.0, 312.0]], "image_paths": ["images/img_000635.png", "images/img_000636.png", "images/img_000629.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000156", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench with armrests'. For each such occluded instance of 'ball', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[453.0, 303.0, 466.0, 322.0], [519.0, 310.0, 527.0, 322.0], [438.0, 303.0, 451.0, 322.0]], "image_paths": ["images/img_000568.png", "images/img_000561.png", "images/img_000571.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000157", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench with armrests'. For each such occluded instance of 'ball', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[411.0, 312.0, 426.0, 332.0]], "image_paths": ["images/img_000612.png", "images/img_000621.png", "images/img_000619.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000158", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stone steps'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[679.0, 107.0, 899.0, 388.0]], "image_paths": ["images/img_000477.png", "images/img_000479.png", "images/img_000471.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000159", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[271.0, 460.0, 336.0, 542.0]], "image_paths": ["images/img_000551.png", "images/img_000546.png", "images/img_000550.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000160", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'natural wooden bench with armrests'. For each such occluded instance of 'ball', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[840.0, 450.0, 851.0, 463.0]], "image_paths": ["images/img_000597.png", "images/img_000603.png", "images/img_000607.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000161", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'white painted wooden bench with armrests'. For each such occluded instance of 'ball', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[525.0, 298.0, 537.0, 315.0]], "image_paths": ["images/img_000595.png", "images/img_000602.png", "images/img_000603.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000162", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench with armrests'. For each such occluded instance of 'ball', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[796.0, 410.0, 816.0, 435.0], [833.0, 432.0, 849.0, 453.0]], "image_paths": ["images/img_000590.png", "images/img_000580.png", "images/img_000586.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000163", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'basketball' that are fully occluded by 'green wooden bench with armrests'. For each such occluded instance of 'basketball', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[752.0, 652.0, 776.0, 683.0], [799.0, 595.0, 821.0, 625.0]], "image_paths": ["images/img_000639.png", "images/img_000629.png", "images/img_000635.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000164", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[274.0, 595.0, 399.0, 710.0]], "image_paths": ["images/img_000542.png", "images/img_000539.png", "images/img_000533.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000165", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[177.0, 228.0, 315.0, 352.0]], "image_paths": ["images/img_000428.png", "images/img_000427.png", "images/img_000420.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000166", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench with armrests'. For each such occluded instance of 'ball', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[774.0, 615.0, 795.0, 642.0], [728.0, 682.0, 741.0, 698.0], [786.0, 602.0, 807.0, 628.0]], "image_paths": ["images/img_000561.png", "images/img_000570.png", "images/img_000571.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000167", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'natural wooden bench with armrests'. For each such occluded instance of 'ball', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[325.0, 312.0, 338.0, 328.0]], "image_paths": ["images/img_000570.png", "images/img_000564.png", "images/img_000572.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000168", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'red shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[438.0, 297.0, 499.0, 345.0]], "image_paths": ["images/img_000518.png", "images/img_000525.png", "images/img_000527.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000169", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[581.0, 363.0, 644.0, 408.0]], "image_paths": ["images/img_000546.png", "images/img_000550.png", "images/img_000554.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000170", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench with armrests'. For each such occluded instance of 'ball', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[453.0, 303.0, 466.0, 322.0], [519.0, 310.0, 527.0, 322.0], [438.0, 303.0, 451.0, 322.0]], "image_paths": ["images/img_000561.png", "images/img_000568.png", "images/img_000571.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000171", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'worker' that are fully occluded by 'stack of sandbags'. For each such occluded instance of 'worker', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[517.0, 258.0, 531.0, 323.0]], "image_paths": ["images/img_000410.png", "images/img_000402.png", "images/img_000408.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000172", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'red shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[739.0, 427.0, 805.0, 492.0]], "image_paths": ["images/img_000518.png", "images/img_000526.png", "images/img_000527.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000173", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'worker' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'worker', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[215.0, 397.0, 242.0, 470.0]], "image_paths": ["images/img_000457.png", "images/img_000448.png", "images/img_000455.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000174", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[454.0, 328.0, 511.0, 393.0]], "image_paths": ["images/img_000546.png", "images/img_000554.png", "images/img_000553.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000175", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stone steps'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[471.0, 73.0, 593.0, 315.0]], "image_paths": ["images/img_000479.png", "images/img_000478.png", "images/img_000471.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000176", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'truck', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[458.0, 172.0, 532.0, 280.0]], "image_paths": ["images/img_000429.png", "images/img_000421.png", "images/img_000428.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000177", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[271.0, 460.0, 336.0, 542.0]], "image_paths": ["images/img_000551.png", "images/img_000550.png", "images/img_000546.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000178", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'stone bench without armrests'. For each such occluded instance of 'ball', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[646.0, 337.0, 661.0, 355.0], [636.0, 337.0, 646.0, 350.0], [672.0, 350.0, 688.0, 370.0], [660.0, 345.0, 674.0, 362.0]], "image_paths": ["images/img_000616.png", "images/img_000622.png", "images/img_000624.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000179", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench'. For each such occluded instance of 'ball', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[220.0, 437.0, 236.0, 457.0], [249.0, 342.0, 264.0, 362.0]], "image_paths": ["images/img_000631.png", "images/img_000638.png", "images/img_000640.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000180", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[119.0, 437.0, 194.0, 513.0]], "image_paths": ["images/img_000490.png", "images/img_000484.png", "images/img_000491.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000181", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench made of blue painted metal'. For each such occluded instance of 'ball', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[818.0, 643.0, 836.0, 668.0], [831.0, 690.0, 845.0, 707.0]], "image_paths": ["images/img_000621.png", "images/img_000617.png", "images/img_000611.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000182", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'basketball' that are fully occluded by 'green wooden bench with armrests'. For each such occluded instance of 'basketball', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[752.0, 652.0, 776.0, 683.0], [799.0, 595.0, 821.0, 625.0]], "image_paths": ["images/img_000639.png", "images/img_000636.png", "images/img_000629.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000183", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'basketball' that are fully occluded by 'stone bench without armrests'. For each such occluded instance of 'basketball', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[725.0, 342.0, 743.0, 365.0]], "image_paths": ["images/img_000586.png", "images/img_000587.png", "images/img_000577.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000184", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stone steps'. For each such occluded instance of 'truck', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[471.0, 73.0, 593.0, 315.0]], "image_paths": ["images/img_000477.png", "images/img_000471.png", "images/img_000478.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000185", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'white shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[158.0, 567.0, 242.0, 617.0]], "image_paths": ["images/img_000496.png", "images/img_000493.png", "images/img_000487.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000186", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[347.0, 367.0, 381.0, 435.0]], "image_paths": ["images/img_000546.png", "images/img_000554.png", "images/img_000552.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000187", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'black stack of bricks'. For each such occluded instance of 'truck', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[0.0, 430.0, 164.0, 555.0]], "image_paths": ["images/img_000455.png", "images/img_000449.png", "images/img_000456.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000188", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'baseball' that are fully occluded by 'blue painted metal bench without armrests'. For each such occluded instance of 'baseball', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[733.0, 348.0, 743.0, 362.0]], "image_paths": ["images/img_000566.png", "images/img_000570.png", "images/img_000560.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000189", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench made of stone'. For each such occluded instance of 'ball', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[236.0, 648.0, 258.0, 675.0], [217.0, 630.0, 230.0, 647.0], [250.0, 665.0, 269.0, 690.0]], "image_paths": ["images/img_000614.png", "images/img_000621.png", "images/img_000620.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000190", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[206.0, 450.0, 321.0, 542.0]], "image_paths": ["images/img_000540.png", "images/img_000543.png", "images/img_000533.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000191", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'basketball' that are fully occluded by 'blue painted metal bench without armrests'. For each such occluded instance of 'basketball', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[135.0, 643.0, 159.0, 673.0], [142.0, 623.0, 166.0, 653.0]], "image_paths": ["images/img_000570.png", "images/img_000566.png", "images/img_000560.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000192", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'white painted wooden bench with armrests'. For each such occluded instance of 'ball', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[525.0, 298.0, 537.0, 315.0]], "image_paths": ["images/img_000595.png", "images/img_000605.png", "images/img_000602.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000193", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[318.0, 582.0, 399.0, 642.0]], "image_paths": ["images/img_000546.png", "images/img_000553.png", "images/img_000550.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000194", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'white shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[491.0, 332.0, 544.0, 393.0]], "image_paths": ["images/img_000549.png", "images/img_000557.png", "images/img_000555.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000195", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'baseball' that are fully occluded by 'stone bench without armrests'. For each such occluded instance of 'baseball', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[453.0, 310.0, 463.0, 323.0]], "image_paths": ["images/img_000624.png", "images/img_000615.png", "images/img_000623.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000196", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench'. For each such occluded instance of 'ball', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[710.0, 375.0, 724.0, 393.0], [835.0, 438.0, 853.0, 460.0]], "image_paths": ["images/img_000641.png", "images/img_000640.png", "images/img_000631.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000197", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[460.0, 168.0, 545.0, 298.0]], "image_paths": ["images/img_000415.png", "images/img_000414.png", "images/img_000406.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000198", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 3 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[111.0, 662.0, 239.0, 738.0]], "image_paths": ["images/img_000483.png", "images/img_000493.png", "images/img_000489.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000199", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'stone bench without armrests'. For each such occluded instance of 'ball', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[636.0, 337.0, 646.0, 350.0], [660.0, 345.0, 674.0, 362.0], [646.0, 337.0, 661.0, 355.0], [672.0, 350.0, 688.0, 370.0], [701.0, 372.0, 710.0, 383.0]], "image_paths": ["images/img_000624.png", "images/img_000615.png", "images/img_000621.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000200", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench made of stone'. For each such occluded instance of 'ball', return its corresponding bounding box in view 1 .\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[689.0, 352.0, 706.0, 375.0], [675.0, 348.0, 689.0, 367.0]], "image_paths": ["images/img_000590.png", "images/img_000588.png", "images/img_000580.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000201", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the rightmost 'street lamp'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[581.0, 248.0, 613.0, 357.0]], "image_paths": ["images/img_000180.png", "images/img_000188.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000202", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the rightmost 'concrete mixer'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[650.0, 317.0, 694.0, 375.0]], "image_paths": ["images/img_000445.png", "images/img_000437.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000203", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the leftmost 'mixed color stack of bricks'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[451.0, 247.0, 545.0, 368.0]], "image_paths": ["images/img_000458.png", "images/img_000450.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000204", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the leftmost 'beige sand helicopter'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[339.0, 343.0, 465.0, 410.0]], "image_paths": ["images/img_000972.png", "images/img_000964.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000205", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the rightmost 'swim ring'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[610.0, 483.0, 661.0, 540.0]], "image_paths": ["images/img_000660.png", "images/img_000669.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000206", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the rightmost 'trolley'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[383.0, 295.0, 432.0, 330.0]], "image_paths": ["images/img_000457.png", "images/img_000450.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000207", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the furthest 'trolley'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[734.0, 343.0, 765.0, 382.0]], "image_paths": ["images/img_000459.png", "images/img_000450.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000208", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the furthest 'green camouflage helicopter'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[667.0, 342.0, 754.0, 398.0]], "image_paths": ["images/img_000980.png", "images/img_000989.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000209", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the nearest 'dark grey helicopter'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[199.0, 403.0, 310.0, 463.0]], "image_paths": ["images/img_001040.png", "images/img_001034.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000210", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the nearest 'stone steps'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[609.0, 325.0, 820.0, 548.0]], "image_paths": ["images/img_000435.png", "images/img_000445.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000211", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the nearest 'airplane with white and green livery'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[665.0, 377.0, 782.0, 467.0]], "image_paths": ["images/img_000231.png", "images/img_000222.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000212", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the leftmost 'F-16 fighter jet'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[159.0, 477.0, 280.0, 590.0]], "image_paths": ["images/img_000890.png", "images/img_000896.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000213", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the nearest 'airplane with yellow livery'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[376.0, 360.0, 509.0, 420.0]], "image_paths": ["images/img_000248.png", "images/img_000254.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000214", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the leftmost 'white tombstone'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[372.0, 393.0, 398.0, 465.0]], "image_paths": ["images/img_000843.png", "images/img_000836.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000215", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the nearest 'blue swim ring'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[559.0, 593.0, 621.0, 662.0]], "image_paths": ["images/img_000704.png", "images/img_000694.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000216", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the nearest 'person saluting'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[610.0, 470.0, 677.0, 608.0]], "image_paths": ["images/img_000686.png", "images/img_000676.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000217", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the nearest 'beige sand helicopter'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[190.0, 397.0, 362.0, 465.0]], "image_paths": ["images/img_000971.png", "images/img_000964.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000218", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the furthest 'basketball'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[226.0, 348.0, 242.0, 370.0]], "image_paths": ["images/img_000627.png", "images/img_000634.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000219", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the furthest 'truck'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[771.0, 453.0, 816.0, 518.0]], "image_paths": ["images/img_000548.png", "images/img_000556.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000220", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the leftmost 'no hat snowman'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[491.0, 252.0, 547.0, 355.0]], "image_paths": ["images/img_000761.png", "images/img_000769.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000221", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the nearest 'F-16 fighter jet'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[554.0, 337.0, 647.0, 407.0]], "image_paths": ["images/img_000890.png", "images/img_000898.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000222", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the leftmost 'stylized F22 fighter jet with white and red color'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[364.0, 337.0, 475.0, 427.0]], "image_paths": ["images/img_000931.png", "images/img_000939.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000223", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the leftmost 'white upright tombstone with cross'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[265.0, 333.0, 311.0, 470.0]], "image_paths": ["images/img_001146.png", "images/img_001153.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000224", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the rightmost 'airplane with light blue livery'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[357.0, 360.0, 491.0, 420.0]], "image_paths": ["images/img_000209.png", "images/img_000215.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000225", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the leftmost 'green camouflage helicopter'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[676.0, 382.0, 771.0, 448.0]], "image_paths": ["images/img_001013.png", "images/img_001023.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000226", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the leftmost 'woman wearing underwear'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[436.0, 335.0, 465.0, 480.0]], "image_paths": ["images/img_001080.png", "images/img_001088.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000227", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the leftmost 'F-22 fighter jet with gray color'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[195.0, 525.0, 330.0, 617.0]], "image_paths": ["images/img_000907.png", "images/img_000902.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000228", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the leftmost 'stack of bricks'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[312.0, 272.0, 380.0, 415.0]], "image_paths": ["images/img_000437.png", "images/img_000444.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000229", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the nearest 'stone steps'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[250.0, 292.0, 409.0, 460.0]], "image_paths": ["images/img_000435.png", "images/img_000442.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000230", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the rightmost 'airplane with yellow livery'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[511.0, 372.0, 645.0, 432.0]], "image_paths": ["images/img_000299.png", "images/img_000291.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000231", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the nearest 'woman in pink clothing'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[464.0, 347.0, 501.0, 470.0]], "image_paths": ["images/img_001112.png", "images/img_001119.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000232", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the nearest 'trellis'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[443.0, 222.0, 453.0, 407.0]], "image_paths": ["images/img_000165.png", "images/img_000159.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000233", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the leftmost 'beach ball'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[438.0, 373.0, 470.0, 415.0]], "image_paths": ["images/img_000646.png", "images/img_000654.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000234", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the nearest 'airplane with light blue livery'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[491.0, 360.0, 624.0, 420.0]], "image_paths": ["images/img_000243.png", "images/img_000235.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000235", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the furthest 'stylized F22 fighter jet with white and red color'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[792.0, 407.0, 931.0, 523.0]], "image_paths": ["images/img_000929.png", "images/img_000919.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000236", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the rightmost 'airplane with light blue livery'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[242.0, 382.0, 376.0, 485.0]], "image_paths": ["images/img_000327.png", "images/img_000320.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000237", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the rightmost 'tombstone with shoulder top'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[546.0, 277.0, 601.0, 357.0]], "image_paths": ["images/img_001133.png", "images/img_001126.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000238", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the rightmost 'bench with armrests'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[566.0, 287.0, 729.0, 400.0]], "image_paths": ["images/img_000589.png", "images/img_000580.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000239", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the nearest 'bench made of blue painted metal'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[235.0, 348.0, 339.0, 550.0]], "image_paths": ["images/img_000583.png", "images/img_000576.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000240", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the nearest 'person'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[333.0, 493.0, 390.0, 647.0]], "image_paths": ["images/img_000666.png", "images/img_000660.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000241", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the leftmost 'blue shipping container'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[551.0, 365.0, 581.0, 433.0]], "image_paths": ["images/img_000498.png", "images/img_000507.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000242", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the leftmost 'woman in gray clothing'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[574.0, 368.0, 611.0, 503.0]], "image_paths": ["images/img_001058.png", "images/img_001049.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000243", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the furthest 'white shipping container'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[667.0, 282.0, 723.0, 383.0]], "image_paths": ["images/img_000106.png", "images/img_000099.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000244", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the furthest 'tombstone'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[379.0, 333.0, 401.0, 405.0]], "image_paths": ["images/img_000806.png", "images/img_000813.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000245", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the furthest 'beige tombstone'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[312.0, 225.0, 349.0, 350.0]], "image_paths": ["images/img_001192.png", "images/img_001199.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000246", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the nearest 'drone with angular white body and gray propellers'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[720.0, 448.0, 781.0, 482.0]], "image_paths": ["images/img_000399.png", "images/img_000391.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000247", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the furthest 'street lamp'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[299.0, 245.0, 335.0, 347.0]], "image_paths": ["images/img_000176.png", "images/img_000170.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000248", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the furthest 'stylized F22 fighter jet with grey and cyan color'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[579.0, 358.0, 637.0, 400.0]], "image_paths": ["images/img_000275.png", "images/img_000266.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000249", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the nearest 'snowman'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[341.0, 505.0, 417.0, 613.0]], "image_paths": ["images/img_000765.png", "images/img_000759.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000250", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the leftmost 'man wearing santa costume'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[611.0, 307.0, 682.0, 502.0]], "image_paths": ["images/img_000068.png", "images/img_000059.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000251", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the nearest 'red street lamp'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[415.0, 228.0, 446.0, 343.0]], "image_paths": ["images/img_000146.png", "images/img_000151.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000252", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the furthest 'woman in pink clothing'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[475.0, 323.0, 532.0, 447.0]], "image_paths": ["images/img_001056.png", "images/img_001049.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000253", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the leftmost 'green shipping container'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[144.0, 515.0, 221.0, 617.0]], "image_paths": ["images/img_000535.png", "images/img_000541.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000254", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the leftmost 'black buttons snowman'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[287.0, 398.0, 369.0, 500.0]], "image_paths": ["images/img_000794.png", "images/img_000800.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000255", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the nearest 'woman in white clothing'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[444.0, 338.0, 471.0, 462.0]], "image_paths": ["images/img_001049.png", "images/img_001055.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000256", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the rightmost 'drone with white oval body shape'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[459.0, 373.0, 497.0, 390.0]], "image_paths": ["images/img_000363.png", "images/img_000367.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000257", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the furthest 'red street lamp'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[671.0, 223.0, 710.0, 333.0]], "image_paths": ["images/img_000152.png", "images/img_000146.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000258", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the leftmost 'man wearing gray windbreaker'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[615.0, 280.0, 649.0, 365.0]], "image_paths": ["images/img_000772.png", "images/img_000782.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000259", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the furthest 'F-16 fighter jet'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[71.0, 458.0, 220.0, 523.0]], "image_paths": ["images/img_000909.png", "images/img_000904.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000260", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the rightmost 'airplane'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[490.0, 373.0, 625.0, 433.0]], "image_paths": ["images/img_000282.png", "images/img_000289.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000261", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the furthest 'olive green helicopter'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[239.0, 358.0, 333.0, 398.0]], "image_paths": ["images/img_001000.png", "images/img_001007.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000262", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the nearest 'bench made of white painted wood'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[281.0, 515.0, 404.0, 667.0]], "image_paths": ["images/img_000603.png", "images/img_000597.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000263", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the leftmost 'soccer ball'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[651.0, 340.0, 666.0, 360.0]], "image_paths": ["images/img_000563.png", "images/img_000573.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000264", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the furthest 'man'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[335.0, 227.0, 375.0, 312.0]], "image_paths": ["images/img_000774.png", "images/img_000781.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000265", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the leftmost 'man wearing santa costume'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[324.0, 267.0, 366.0, 368.0]], "image_paths": ["images/img_000726.png", "images/img_000734.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000266", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the nearest 'harrier jumpjet fighter jet'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[168.0, 495.0, 279.0, 587.0]], "image_paths": ["images/img_000937.png", "images/img_000943.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000267", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the leftmost 'stylized F22 fighter jet with white and red color'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[339.0, 327.0, 454.0, 418.0]], "image_paths": ["images/img_000911.png", "images/img_000904.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000268", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the nearest 'green shipping container'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[709.0, 423.0, 784.0, 558.0]], "image_paths": ["images/img_000110.png", "images/img_000116.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000269", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the nearest 'man wearing black coat'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[484.0, 258.0, 527.0, 457.0]], "image_paths": ["images/img_000004.png", "images/img_000013.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000270", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the furthest 'red shipping container'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[237.0, 363.0, 334.0, 433.0]], "image_paths": ["images/img_000092.png", "images/img_000088.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000271", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the leftmost 'red gloves snowman'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[470.0, 222.0, 541.0, 332.0]], "image_paths": ["images/img_000782.png", "images/img_000774.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000272", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the leftmost 'stone steps'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[125.0, 353.0, 340.0, 580.0]], "image_paths": ["images/img_000467.png", "images/img_000473.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000273", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the furthest 'beach ball'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[843.0, 427.0, 859.0, 447.0]], "image_paths": ["images/img_000603.png", "images/img_000593.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000274", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the leftmost 'black upright tombstone with cross'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[376.0, 243.0, 417.0, 368.0]], "image_paths": ["images/img_001166.png", "images/img_001159.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000275", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the furthest 'green street lamp'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[333.0, 235.0, 355.0, 342.0]], "image_paths": ["images/img_000135.png", "images/img_000139.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000276", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the furthest 'drone with white oval body shape'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[812.0, 480.0, 846.0, 502.0]], "image_paths": ["images/img_000351.png", "images/img_000358.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000277", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the rightmost 'swim ring'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[352.0, 397.0, 404.0, 450.0]], "image_paths": ["images/img_000718.png", "images/img_000712.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000278", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the leftmost 'red shipping container'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[198.0, 450.0, 311.0, 532.0]], "image_paths": ["images/img_000116.png", "images/img_000112.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000279", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the nearest 'airplane with white and green livery'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[341.0, 348.0, 444.0, 428.0]], "image_paths": ["images/img_000222.png", "images/img_000229.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000280", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the leftmost 'stylized F22 fighter jet with camoflague color'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[339.0, 345.0, 449.0, 430.0]], "image_paths": ["images/img_000957.png", "images/img_000949.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000281", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the rightmost 'man wearing green hood'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[526.0, 243.0, 578.0, 448.0]], "image_paths": ["images/img_000049.png", "images/img_000057.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000282", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the nearest 'lamp'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[626.0, 302.0, 655.0, 348.0]], "image_paths": ["images/img_000475.png", "images/img_000465.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000283", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the rightmost 'tombstone with shoulder top'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[669.0, 510.0, 741.0, 600.0]], "image_paths": ["images/img_001135.png", "images/img_001126.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000284", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the leftmost 'woman wearing scarf accessory'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[424.0, 343.0, 458.0, 480.0]], "image_paths": ["images/img_001122.png", "images/img_001114.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000285", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the furthest 'worker'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[856.0, 403.0, 887.0, 483.0]], "image_paths": ["images/img_000425.png", "images/img_000416.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000286", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the leftmost 'beach ball'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[339.0, 430.0, 375.0, 478.0]], "image_paths": ["images/img_000713.png", "images/img_000707.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000287", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the nearest 'red shipping container'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[256.0, 547.0, 364.0, 692.0]], "image_paths": ["images/img_000088.png", "images/img_000091.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000288", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the rightmost 'woman in gray clothing'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[394.0, 372.0, 439.0, 508.0]], "image_paths": ["images/img_001072.png", "images/img_001066.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000289", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the leftmost 'man wearing santa costume'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[530.0, 263.0, 581.0, 448.0]], "image_paths": ["images/img_000059.png", "images/img_000067.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000290", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the leftmost 'white tombstone'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[549.0, 255.0, 608.0, 353.0]], "image_paths": ["images/img_001186.png", "images/img_001177.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000291", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the nearest 'white tombstone'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[549.0, 255.0, 608.0, 353.0]], "image_paths": ["images/img_001177.png", "images/img_001186.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000292", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the rightmost 'shipping container'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[719.0, 483.0, 820.0, 615.0]], "image_paths": ["images/img_000085.png", "images/img_000076.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000293", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the rightmost 'man wearing gray windbreaker'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[610.0, 303.0, 676.0, 498.0]], "image_paths": ["images/img_000013.png", "images/img_000004.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000294", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the nearest 'green shipping container'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[481.0, 285.0, 520.0, 385.0]], "image_paths": ["images/img_000080.png", "images/img_000074.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000295", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the leftmost 'drone with rectangular white and black body'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[359.0, 315.0, 431.0, 355.0]], "image_paths": ["images/img_000388.png", "images/img_000394.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000296", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the leftmost 'F-22 fighter jet with gray color'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[145.0, 483.0, 279.0, 592.0]], "image_paths": ["images/img_000959.png", "images/img_000953.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000297", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the nearest 'woman in pink clothing'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[505.0, 352.0, 564.0, 475.0]], "image_paths": ["images/img_001120.png", "images/img_001112.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000298", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the leftmost 'shipping container'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[562.0, 300.0, 646.0, 365.0]], "image_paths": ["images/img_000076.png", "images/img_000084.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000299", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the rightmost 'man in standing posture'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[540.0, 308.0, 564.0, 408.0]], "image_paths": ["images/img_000806.png", "images/img_000814.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000300", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the leftmost 'truck with white and black cab'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[201.0, 603.0, 306.0, 750.0]], "image_paths": ["images/img_000500.png", "images/img_000506.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000301", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the nearest 'drone with white oval body shape'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[188.0, 562.0, 226.0, 592.0]], "image_paths": ["images/img_000348.png", "images/img_000342.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000302", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the rightmost 'airplane with yellow livery'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[205.0, 413.0, 344.0, 462.0]], "image_paths": ["images/img_000222.png", "images/img_000227.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000303", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the nearest 'airplane with light blue livery'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[193.0, 420.0, 336.0, 478.0]], "image_paths": ["images/img_000235.png", "images/img_000241.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000304", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the furthest 'drone with white oval body shape'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[249.0, 383.0, 274.0, 400.0]], "image_paths": ["images/img_000377.png", "images/img_000382.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000305", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the rightmost 'watering can'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[389.0, 290.0, 444.0, 357.0]], "image_paths": ["images/img_000175.png", "images/img_000170.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000306", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the rightmost 'green shipping container'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[594.0, 480.0, 677.0, 538.0]], "image_paths": ["images/img_000540.png", "images/img_000531.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000307", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the furthest 'man in crouching posture'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[639.0, 457.0, 667.0, 513.0]], "image_paths": ["images/img_000824.png", "images/img_000834.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000308", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the leftmost 'red gloves snowman'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[310.0, 418.0, 356.0, 505.0]], "image_paths": ["images/img_000790.png", "images/img_000796.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000309", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the furthest 'stylized F22 fighter jet with camoflague color'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[789.0, 428.0, 865.0, 472.0]], "image_paths": ["images/img_000308.png", "images/img_000316.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000310", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the rightmost 'olive green helicopter'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[381.0, 355.0, 477.0, 395.0]], "image_paths": ["images/img_001013.png", "images/img_001020.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000311", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the rightmost 'drone with white oval body shape'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[482.0, 398.0, 520.0, 412.0]], "image_paths": ["images/img_000351.png", "images/img_000356.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000312", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the leftmost 'airplane with white and green livery'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[665.0, 377.0, 782.0, 467.0]], "image_paths": ["images/img_000222.png", "images/img_000231.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000313", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the furthest 'green shipping container'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[463.0, 240.0, 551.0, 333.0]], "image_paths": ["images/img_000131.png", "images/img_000124.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000314", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the rightmost 'man wearing red jacket'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[590.0, 285.0, 650.0, 478.0]], "image_paths": ["images/img_000023.png", "images/img_000016.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000315", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the nearest 'beach ball'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[499.0, 348.0, 531.0, 392.0]], "image_paths": ["images/img_000715.png", "images/img_000707.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000316", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the leftmost 'person'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[505.0, 238.0, 554.0, 373.0]], "image_paths": ["images/img_000703.png", "images/img_000694.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000317", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the furthest 'man wearing gray hat'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[345.0, 265.0, 388.0, 458.0]], "image_paths": ["images/img_000040.png", "images/img_000033.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000318", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the leftmost 'black tombstone'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[378.0, 287.0, 492.0, 372.0]], "image_paths": ["images/img_001165.png", "images/img_001157.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000319", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the leftmost 'lamp'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[626.0, 302.0, 655.0, 348.0]], "image_paths": ["images/img_000475.png", "images/img_000465.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000320", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the nearest 'airplane with red livery'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[359.0, 372.0, 492.0, 432.0]], "image_paths": ["images/img_000266.png", "images/img_000275.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000321", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the leftmost 'F-16 fighter jet'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[665.0, 373.0, 765.0, 455.0]], "image_paths": ["images/img_000945.png", "images/img_000935.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000322", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the leftmost 'woman wearing black dress'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[395.0, 400.0, 444.0, 553.0]], "image_paths": ["images/img_001095.png", "images/img_001101.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000323", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the leftmost 'pink flamingo swim ring'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[329.0, 540.0, 388.0, 625.0]], "image_paths": ["images/img_000686.png", "images/img_000680.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000324", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the leftmost 'stack of bricks'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[249.0, 380.0, 336.0, 558.0]], "image_paths": ["images/img_000426.png", "images/img_000419.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000325", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the nearest 'airplane with yellow livery'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[233.0, 423.0, 371.0, 480.0]], "image_paths": ["images/img_000294.png", "images/img_000301.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000326", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the leftmost 'person wearing blue patterned shirt'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[345.0, 253.0, 380.0, 407.0]], "image_paths": ["images/img_000676.png", "images/img_000683.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000327", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the rightmost 'black tombstone'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[519.0, 312.0, 639.0, 430.0]], "image_paths": ["images/img_001157.png", "images/img_001164.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000328", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the rightmost 'F-22 fighter jet with gray color'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[520.0, 373.0, 630.0, 427.0]], "image_paths": ["images/img_000949.png", "images/img_000957.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000329", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the nearest 'airplane'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[228.0, 425.0, 365.0, 473.0]], "image_paths": ["images/img_000282.png", "images/img_000288.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000330", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the leftmost 'drone with angular black body'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[286.0, 367.0, 388.0, 430.0]], "image_paths": ["images/img_000376.png", "images/img_000381.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000331", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the leftmost 'woman in white clothing'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[511.0, 345.0, 557.0, 468.0]], "image_paths": ["images/img_001060.png", "images/img_001069.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000332", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the furthest 'airplane with white and green livery'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[242.0, 360.0, 375.0, 420.0]], "image_paths": ["images/img_000206.png", "images/img_000211.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000333", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the rightmost 'two balls snowman'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[490.0, 282.0, 526.0, 378.0]], "image_paths": ["images/img_000745.png", "images/img_000752.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000334", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the rightmost 'black street lamp'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[560.0, 243.0, 590.0, 350.0]], "image_paths": ["images/img_000157.png", "images/img_000161.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000335", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the leftmost 'man wearing green hood'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[629.0, 290.0, 686.0, 512.0]], "image_paths": ["images/img_000057.png", "images/img_000047.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000336", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the furthest 'airplane with yellow livery'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[244.0, 360.0, 376.0, 420.0]], "image_paths": ["images/img_000248.png", "images/img_000254.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000337", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the leftmost 'stylized F22 fighter jet with camoflague color'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[147.0, 468.0, 289.0, 597.0]], "image_paths": ["images/img_000921.png", "images/img_000927.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000338", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the leftmost 'snowman'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[274.0, 395.0, 361.0, 512.0]], "image_paths": ["images/img_000747.png", "images/img_000741.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000339", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the leftmost 'F-22 fighter jet with gray color'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[177.0, 415.0, 260.0, 445.0]], "image_paths": ["images/img_000309.png", "images/img_000317.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000340", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the leftmost 'beach ball'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[250.0, 487.0, 266.0, 507.0]], "image_paths": ["images/img_000619.png", "images/img_000612.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000341", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the leftmost 'drone with smooth black curved body'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[665.0, 417.0, 746.0, 462.0]], "image_paths": ["images/img_000348.png", "images/img_000338.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000342", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the furthest 'olive green helicopter'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[465.0, 315.0, 526.0, 365.0]], "image_paths": ["images/img_001008.png", "images/img_001000.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000343", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the leftmost 'black buttons snowman'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[624.0, 305.0, 650.0, 382.0]], "image_paths": ["images/img_000733.png", "images/img_000724.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000344", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the leftmost 'black street lamp'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[654.0, 328.0, 684.0, 438.0]], "image_paths": ["images/img_000142.png", "images/img_000135.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000345", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the leftmost 'beach ball'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[190.0, 508.0, 207.0, 530.0]], "image_paths": ["images/img_000600.png", "images/img_000593.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000346", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the furthest 'green camouflage helicopter'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[667.0, 342.0, 754.0, 398.0]], "image_paths": ["images/img_000989.png", "images/img_000980.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000347", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the furthest 'airplane with white and green livery'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[46.0, 453.0, 215.0, 508.0]], "image_paths": ["images/img_000257.png", "images/img_000251.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000348", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the furthest 'man wearing red jacket'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[481.0, 240.0, 522.0, 423.0]], "image_paths": ["images/img_000066.png", "images/img_000059.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000349", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the nearest 'green street lamp'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[269.0, 357.0, 309.0, 462.0]], "image_paths": ["images/img_000157.png", "images/img_000160.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000350", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the furthest 'worker'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[664.0, 255.0, 686.0, 318.0]], "image_paths": ["images/img_000419.png", "images/img_000428.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000351", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the furthest 'truck'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[468.0, 292.0, 529.0, 328.0]], "image_paths": ["images/img_000548.png", "images/img_000554.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000352", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the furthest 'truck'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[321.0, 345.0, 376.0, 395.0]], "image_paths": ["images/img_000482.png", "images/img_000490.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000353", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the furthest 'beige tombstone'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[389.0, 358.0, 405.0, 408.0]], "image_paths": ["images/img_000843.png", "images/img_000836.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000354", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the leftmost 'tombstone'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[356.0, 402.0, 385.0, 480.0]], "image_paths": ["images/img_000878.png", "images/img_000871.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000355", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the nearest 'bench with armrests'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[304.0, 285.0, 459.0, 447.0]], "image_paths": ["images/img_000627.png", "images/img_000635.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000356", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the leftmost 'tombstone'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[517.0, 317.0, 541.0, 390.0]], "image_paths": ["images/img_000854.png", "images/img_000863.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000357", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the furthest 'stylized F22 fighter jet with grey and cyan color'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[371.0, 355.0, 429.0, 395.0]], "image_paths": ["images/img_000320.png", "images/img_000328.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000358", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the rightmost 'beige tombstone'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[696.0, 323.0, 740.0, 455.0]], "image_paths": ["images/img_001177.png", "images/img_001186.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000359", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the leftmost 'white shipping container'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[240.0, 413.0, 311.0, 548.0]], "image_paths": ["images/img_000114.png", "images/img_000110.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000360", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the nearest 'green shipping container'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[291.0, 483.0, 390.0, 593.0]], "image_paths": ["images/img_000492.png", "images/img_000486.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000361", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the rightmost 'man in sitting posture'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[441.0, 355.0, 465.0, 415.0]], "image_paths": ["images/img_000861.png", "images/img_000854.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000362", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the rightmost 'drone with smooth black curved body'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[566.0, 402.0, 641.0, 435.0]], "image_paths": ["images/img_000400.png", "images/img_000392.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000363", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the nearest 'flat tombstone'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[237.0, 518.0, 395.0, 690.0]], "image_paths": ["images/img_001198.png", "images/img_001192.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000364", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the leftmost 'man wearing green hood'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[315.0, 347.0, 374.0, 585.0]], "image_paths": ["images/img_000053.png", "images/img_000047.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000365", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the furthest 'black tombstone'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[453.0, 165.0, 494.0, 293.0]], "image_paths": ["images/img_001146.png", "images/img_001154.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000366", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the nearest 'watering can'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[639.0, 358.0, 700.0, 437.0]], "image_paths": ["images/img_000170.png", "images/img_000177.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000367", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the furthest 'desert camouflage helicopter'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[241.0, 343.0, 339.0, 393.0]], "image_paths": ["images/img_001013.png", "images/img_001020.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000368", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the rightmost 'airplane with light blue livery'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[465.0, 573.0, 644.0, 670.0]], "image_paths": ["images/img_000317.png", "images/img_000308.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000369", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the furthest 'upright tombstone'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[380.0, 363.0, 417.0, 400.0]], "image_paths": ["images/img_000858.png", "images/img_000865.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000370", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the leftmost 'tombstone'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[356.0, 402.0, 385.0, 480.0]], "image_paths": ["images/img_000871.png", "images/img_000878.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000371", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the nearest 'man wearing brown coat'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[354.0, 413.0, 405.0, 613.0]], "image_paths": ["images/img_000012.png", "images/img_000006.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000372", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the rightmost 'blue shipping container'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[652.0, 423.0, 731.0, 520.0]], "image_paths": ["images/img_000482.png", "images/img_000492.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000373", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the furthest 'man wearing brown coat'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[323.0, 265.0, 365.0, 457.0]], "image_paths": ["images/img_000042.png", "images/img_000035.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000374", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the nearest 'black hat snowman'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[490.0, 282.0, 526.0, 378.0]], "image_paths": ["images/img_000752.png", "images/img_000745.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000375", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the furthest 'basketball'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[823.0, 433.0, 841.0, 457.0]], "image_paths": ["images/img_000573.png", "images/img_000563.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000376", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the rightmost 'woman wearing underwear'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[464.0, 332.0, 494.0, 470.0]], "image_paths": ["images/img_001102.png", "images/img_001095.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000377", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the rightmost 'red shipping container'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[484.0, 288.0, 519.0, 380.0]], "image_paths": ["images/img_000098.png", "images/img_000102.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000378", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the nearest 'green shipping container'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[585.0, 398.0, 655.0, 480.0]], "image_paths": ["images/img_000523.png", "images/img_000514.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000379", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the rightmost 'green shipping container'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[522.0, 357.0, 552.0, 433.0]], "image_paths": ["images/img_000522.png", "images/img_000516.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000380", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the rightmost 'bench made of red painted metal'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[618.0, 325.0, 723.0, 373.0]], "image_paths": ["images/img_000636.png", "images/img_000627.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000381", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the leftmost 'woman in white clothing'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[511.0, 345.0, 557.0, 468.0]], "image_paths": ["images/img_001069.png", "images/img_001060.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000382", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the rightmost 'drone with smooth black curved body'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[787.0, 515.0, 890.0, 565.0]], "image_paths": ["images/img_000371.png", "images/img_000364.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000383", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the furthest 'trellis'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[472.0, 167.0, 555.0, 320.0]], "image_paths": ["images/img_000180.png", "images/img_000188.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000384", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the nearest 'green shipping container'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[546.0, 392.0, 610.0, 467.0]], "image_paths": ["images/img_000504.png", "images/img_000512.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000385", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the rightmost 'man wearing gray windbreaker'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[586.0, 402.0, 646.0, 633.0]], "image_paths": ["images/img_000019.png", "images/img_000028.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000386", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the nearest 'green shipping container'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[611.0, 545.0, 756.0, 712.0]], "image_paths": ["images/img_000133.png", "images/img_000124.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000387", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the furthest 'olive green helicopter'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[104.0, 442.0, 163.0, 505.0]], "image_paths": ["images/img_001032.png", "images/img_001037.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000388", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the nearest 'bench made of stone'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[448.0, 308.0, 536.0, 387.0]], "image_paths": ["images/img_000623.png", "images/img_000614.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000389", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the nearest 'person'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[557.0, 318.0, 604.0, 470.0]], "image_paths": ["images/img_000652.png", "images/img_000644.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000390", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the furthest 'airplane with white and green livery'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[641.0, 373.0, 765.0, 418.0]], "image_paths": ["images/img_000233.png", "images/img_000241.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000391", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the furthest 'man wearing gray hat'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[662.0, 273.0, 694.0, 365.0]], "image_paths": ["images/img_000757.png", "images/img_000766.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000392", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the leftmost 'man in standing posture'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[583.0, 368.0, 608.0, 463.0]], "image_paths": ["images/img_000834.png", "images/img_000824.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000393", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the rightmost 'man in sitting posture'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[575.0, 522.0, 618.0, 578.0]], "image_paths": ["images/img_000824.png", "images/img_000834.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000394", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the leftmost 'woman wearing underwear'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[561.0, 350.0, 603.0, 498.0]], "image_paths": ["images/img_001080.png", "images/img_001090.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000395", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the leftmost 'bench made of red painted metal'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[172.0, 415.0, 290.0, 523.0]], "image_paths": ["images/img_000582.png", "images/img_000588.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000396", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the leftmost 'dark grey helicopter'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[671.0, 385.0, 766.0, 453.0]], "image_paths": ["images/img_001043.png", "images/img_001034.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000397", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the leftmost 'blue shipping container'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[451.0, 300.0, 544.0, 368.0]], "image_paths": ["images/img_000124.png", "images/img_000132.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000398", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the nearest 'airplane with white and orange livery'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[376.0, 355.0, 490.0, 435.0]], "image_paths": ["images/img_000277.png", "images/img_000284.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000399", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the nearest 'airplane with red livery'. \nThen return the bounding box of the target instance in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[124.0, 503.0, 304.0, 572.0]], "image_paths": ["images/img_000214.png", "images/img_000209.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000400", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the nearest 'drone with rectangular white and black body'. \nThen return the bounding box of the target instance in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[264.0, 388.0, 340.0, 430.0]], "image_paths": ["images/img_000340.png", "images/img_000348.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000401", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('man wearing gray hat') is boxed.\nIn view 2, locate all target instances 'coupe' that lie in the direct-left direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[443.0, 30.0, 536.0, 252.0]], "image_paths": ["images/img_000738.png", "images/img_000746.png", "images/img_000752.png"], "supporting": {"draw_boxes": {"images/img_000738.png": [[374, 126, 63, 208]]}}} +{"q_idx": "detect_000402", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('tombstone with pointed top') is boxed.\nIn view 3, locate all target instances 'tombstone with pointed top' that lie in the direct-right direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[263.0, 482.0, 331.0, 612.0]], "image_paths": ["images/img_001174.png", "images/img_001187.png", "images/img_001183.png"], "supporting": {"draw_boxes": {"images/img_001174.png": [[221, 193, 97, 183]]}}} +{"q_idx": "detect_000403", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('airplane with red livery') is boxed.\nIn view 1, locate all target instances 'airplane with red livery' that lie in the right-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[206.0, 380.0, 344.0, 487.0], [156.0, 432.0, 318.0, 565.0]], "image_paths": ["images/img_000202.png", "images/img_000193.png", "images/img_000200.png"], "supporting": {"draw_boxes": {"images/img_000193.png": [[0, 190, 110, 46]]}}} +{"q_idx": "detect_000404", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('person saluting') is boxed.\nIn view 3, locate all target instances 'toy crab' that lie in the direct-right direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[510.0, 352.0, 551.0, 397.0]], "image_paths": ["images/img_000685.png", "images/img_000679.png", "images/img_000687.png"], "supporting": {"draw_boxes": {"images/img_000679.png": [[234, 63, 148, 343]]}}} +{"q_idx": "detect_000405", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('woman in gray clothing') is boxed.\nIn view 2, locate all target instances 'woman in white clothing' that lie in the direct-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[606.0, 402.0, 650.0, 530.0]], "image_paths": ["images/img_001063.png", "images/img_001072.png"], "supporting": {"draw_boxes": {"images/img_001063.png": [[0, 34, 120, 429]]}}} +{"q_idx": "detect_000406", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('plaid scarf snowman') is boxed.\nIn view 1, locate all target instances 'man wearing red jacket' that lie in the direct-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[247.0, 347.0, 285.0, 447.0]], "image_paths": ["images/img_000730.png", "images/img_000724.png"], "supporting": {"draw_boxes": {"images/img_000724.png": [[302, 154, 92, 186]]}}} +{"q_idx": "detect_000407", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('bench made of red painted metal') is boxed.\nIn view 2, locate all target instances 'soccer ball' that lie in the left-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[854.0, 663.0, 876.0, 692.0]], "image_paths": ["images/img_000628.png", "images/img_000634.png"], "supporting": {"draw_boxes": {"images/img_000628.png": [[649, 291, 151, 94]]}}} +{"q_idx": "detect_000408", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('beige sand helicopter') is boxed.\nIn view 2, locate all target instances 'olive green helicopter' that lie in the left-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[64.0, 445.0, 177.0, 507.0]], "image_paths": ["images/img_000971.png", "images/img_000974.png", "images/img_000965.png"], "supporting": {"draw_boxes": {"images/img_000965.png": [[442, 159, 190, 83]]}}} +{"q_idx": "detect_000409", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('man wearing santa costume') is boxed.\nIn view 1, locate all target instances 'man wearing black coat' that lie in the direct-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[258.0, 335.0, 314.0, 557.0]], "image_paths": ["images/img_000055.png", "images/img_000053.png", "images/img_000046.png"], "supporting": {"draw_boxes": {"images/img_000046.png": [[0, 13, 121, 436]]}}} +{"q_idx": "detect_000410", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('man in crouching posture') is boxed.\nIn view 1, locate all target instances 'tombstone without cross' that lie in the direct-right direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[596.0, 395.0, 636.0, 465.0]], "image_paths": ["images/img_000830.png", "images/img_000828.png", "images/img_000821.png"], "supporting": {"draw_boxes": {"images/img_000821.png": [[0, 209, 76, 160]]}}} +{"q_idx": "detect_000411", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('airplane with white and green livery') is boxed.\nIn view 2, locate all target instances 'airplane with red livery' that lie in the direct-right direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[245.0, 512.0, 410.0, 658.0]], "image_paths": ["images/img_000197.png", "images/img_000201.png", "images/img_000193.png"], "supporting": {"draw_boxes": {"images/img_000193.png": [[382, 148, 274, 89]]}}} +{"q_idx": "detect_000412", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('man wearing a green hoodie and orange shoes') is boxed.\nIn view 1, locate all target instances 'man wearing a green hoodie and orange shoes' that lie in the direct-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[346.0, 437.0, 376.0, 533.0]], "image_paths": ["images/img_000863.png", "images/img_000866.png", "images/img_000853.png"], "supporting": {"draw_boxes": {"images/img_000853.png": [[462, 59, 131, 314]]}}} +{"q_idx": "detect_000413", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('airplane with white and orange livery') is boxed.\nIn view 1, locate all target instances 'airplane with white and green livery' that lie in the left-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[427.0, 650.0, 634.0, 777.0], [580.0, 567.0, 774.0, 657.0]], "image_paths": ["images/img_000225.png", "images/img_000219.png"], "supporting": {"draw_boxes": {"images/img_000219.png": [[0, 148, 163, 88]]}}} +{"q_idx": "detect_000414", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('airplane with yellow livery') is boxed.\nIn view 2, locate all target instances 'mirage fighter jet' that lie in the left-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[163.0, 427.0, 240.0, 455.0]], "image_paths": ["images/img_000326.png", "images/img_000332.png"], "supporting": {"draw_boxes": {"images/img_000326.png": [[672, 109, 128, 103]]}}} +{"q_idx": "detect_000415", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('tombstone without cross') is boxed.\nIn view 1, locate all target instances 'man in crouching posture' that lie in the direct-left direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[370.0, 407.0, 401.0, 468.0]], "image_paths": ["images/img_000813.png", "images/img_000806.png"], "supporting": {"draw_boxes": {"images/img_000806.png": [[345, 203, 92, 128]]}}} +{"q_idx": "detect_000416", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('stylized F22 fighter jet with grey and cyan color') is boxed.\nIn view 2, locate all target instances 'mirage fighter jet' that lie in the direct-left direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[641.0, 663.0, 780.0, 727.0]], "image_paths": ["images/img_000320.png", "images/img_000329.png"], "supporting": {"draw_boxes": {"images/img_000320.png": [[253, 198, 96, 38]]}}} +{"q_idx": "detect_000417", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('stylized F22 fighter jet with grey and cyan color') is boxed.\nIn view 1, locate all target instances 'F-16 fighter jet' that lie in the direct-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[664.0, 377.0, 730.0, 417.0]], "image_paths": ["images/img_000319.png", "images/img_000310.png"], "supporting": {"draw_boxes": {"images/img_000310.png": [[520, 195, 100, 41]]}}} +{"q_idx": "detect_000418", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('stylized F22 fighter jet with camoflague color') is boxed.\nIn view 1, locate all target instances 'airplane with white and orange livery' that lie in the right-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[356.0, 372.0, 490.0, 432.0]], "image_paths": ["images/img_000303.png", "images/img_000302.png", "images/img_000296.png"], "supporting": {"draw_boxes": {"images/img_000296.png": [[722, 181, 78, 57]]}}} +{"q_idx": "detect_000419", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('bench made of black metal') is boxed.\nIn view 2, locate all target instances 'bench made of black metal' that lie in the direct-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[674.0, 493.0, 804.0, 732.0]], "image_paths": ["images/img_000637.png", "images/img_000634.png", "images/img_000628.png"], "supporting": {"draw_boxes": {"images/img_000628.png": [[0, 222, 204, 179]]}}} +{"q_idx": "detect_000420", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('airplane with red livery') is boxed.\nIn view 2, locate all target instances 'airplane with yellow livery' that lie in the left-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[644.0, 405.0, 796.0, 473.0]], "image_paths": ["images/img_000254.png", "images/img_000258.png", "images/img_000252.png"], "supporting": {"draw_boxes": {"images/img_000252.png": [[160, 169, 229, 69]]}}} +{"q_idx": "detect_000421", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('person with black hat') is boxed.\nIn view 3, locate all target instances 'person saluting' that lie in the direct-left direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[411.0, 267.0, 444.0, 398.0]], "image_paths": ["images/img_000704.png", "images/img_000695.png", "images/img_000702.png"], "supporting": {"draw_boxes": {"images/img_000695.png": [[91, 40, 126, 367]]}}} +{"q_idx": "detect_000422", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('airplane with white and orange livery') is boxed.\nIn view 2, locate all target instances 'F-16 fighter jet' that lie in the direct-right direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[165.0, 595.0, 247.0, 673.0]], "image_paths": ["images/img_000282.png", "images/img_000290.png"], "supporting": {"draw_boxes": {"images/img_000282.png": [[680, 108, 120, 103]]}}} +{"q_idx": "detect_000423", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('beige upright tombstone with cross') is boxed.\nIn view 3, locate all target instances 'tombstone with pointed top' that lie in the right-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[186.0, 390.0, 263.0, 513.0]], "image_paths": ["images/img_001127.png", "images/img_001136.png", "images/img_001135.png"], "supporting": {"draw_boxes": {"images/img_001127.png": [[493, 129, 87, 235]]}}} +{"q_idx": "detect_000424", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('person with black hat') is boxed.\nIn view 1, locate all target instances 'lifeguard tower' that lie in the left-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[405.0, 2.0, 986.0, 390.0]], "image_paths": ["images/img_000704.png", "images/img_000698.png", "images/img_000695.png"], "supporting": {"draw_boxes": {"images/img_000695.png": [[91, 40, 126, 367]]}}} +{"q_idx": "detect_000425", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('concrete mixer') is boxed.\nIn view 2, locate all target instances 'concrete mixer' that lie in the right-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[838.0, 417.0, 892.0, 482.0], [802.0, 395.0, 849.0, 463.0]], "image_paths": ["images/img_000435.png", "images/img_000441.png", "images/img_000445.png"], "supporting": {"draw_boxes": {"images/img_000435.png": [[48, 207, 103, 84]]}}} +{"q_idx": "detect_000426", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('man wearing green hood') is boxed.\nIn view 1, locate all target instances 'man wearing black coat' that lie in the direct-right direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[465.0, 423.0, 535.0, 673.0]], "image_paths": ["images/img_000027.png", "images/img_000019.png"], "supporting": {"draw_boxes": {"images/img_000019.png": [[0, 0, 158, 552]]}}} +{"q_idx": "detect_000427", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('green wooden bench with armrests') is boxed.\nIn view 2, locate all target instances 'basketball' that lie in the left-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[255.0, 385.0, 273.0, 407.0], [400.0, 308.0, 415.0, 328.0]], "image_paths": ["images/img_000615.png", "images/img_000622.png"], "supporting": {"draw_boxes": {"images/img_000615.png": [[0, 229, 221, 157]]}}} +{"q_idx": "detect_000428", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('man wearing santa costume') is boxed.\nIn view 1, locate all target instances 'taxi' that lie in the left-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[181.0, 133.0, 326.0, 290.0]], "image_paths": ["images/img_000802.png", "images/img_000794.png", "images/img_000796.png"], "supporting": {"draw_boxes": {"images/img_000794.png": [[473, 140, 81, 175]]}}} +{"q_idx": "detect_000429", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('woman in pink clothing') is boxed.\nIn view 2, locate all target instances 'woman in white clothing' that lie in the direct-right direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[547.0, 382.0, 596.0, 508.0]], "image_paths": ["images/img_001049.png", "images/img_001057.png"], "supporting": {"draw_boxes": {"images/img_001049.png": [[312, 79, 180, 327]]}}} +{"q_idx": "detect_000430", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('bench made of blue painted metal') is boxed.\nIn view 1, locate all target instances 'bench made of black metal' that lie in the direct-left direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[665.0, 507.0, 791.0, 747.0]], "image_paths": ["images/img_000570.png", "images/img_000568.png", "images/img_000560.png"], "supporting": {"draw_boxes": {"images/img_000560.png": [[0, 227, 233, 169]]}}} +{"q_idx": "detect_000431", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('toy crab') is boxed.\nIn view 1, locate all target instances 'beach ball' that lie in the left-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[652.0, 623.0, 694.0, 678.0]], "image_paths": ["images/img_000719.png", "images/img_000715.png", "images/img_000712.png"], "supporting": {"draw_boxes": {"images/img_000712.png": [[244, 347, 117, 40]]}}} +{"q_idx": "detect_000432", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('man wearing gray windbreaker') is boxed.\nIn view 2, locate all target instances 'man wearing black coat' that lie in the direct-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[618.0, 298.0, 677.0, 512.0]], "image_paths": ["images/img_000020.png", "images/img_000028.png"], "supporting": {"draw_boxes": {"images/img_000020.png": [[121, 0, 160, 475]]}}} +{"q_idx": "detect_000433", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('surfboard') is boxed.\nIn view 1, locate all target instances 'beach ball' that lie in the direct-right direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[615.0, 425.0, 650.0, 470.0]], "image_paths": ["images/img_000669.png", "images/img_000668.png", "images/img_000663.png"], "supporting": {"draw_boxes": {"images/img_000663.png": [[618, 0, 156, 399]]}}} +{"q_idx": "detect_000434", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('snowman with black hat and red scarf') is boxed.\nIn view 1, locate all target instances 'snowman with black hat and red gloves' that lie in the direct-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[286.0, 408.0, 360.0, 513.0]], "image_paths": ["images/img_000749.png", "images/img_000740.png", "images/img_000752.png"], "supporting": {"draw_boxes": {"images/img_000740.png": [[0, 46, 216, 331]]}}} +{"q_idx": "detect_000435", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('toy crab') is boxed.\nIn view 3, locate all target instances 'tree' that lie in the left-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[0.0, 0.0, 164.0, 527.0]], "image_paths": ["images/img_000720.png", "images/img_000712.png", "images/img_000719.png"], "supporting": {"draw_boxes": {"images/img_000712.png": [[244, 347, 117, 40]]}}} +{"q_idx": "detect_000436", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('tombstone without cross') is boxed.\nIn view 3, locate all target instances 'man wearing a green hoodie and orange shoes' that lie in the direct-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[454.0, 297.0, 489.0, 395.0]], "image_paths": ["images/img_000805.png", "images/img_000816.png", "images/img_000813.png"], "supporting": {"draw_boxes": {"images/img_000805.png": [[0, 220, 89, 140]]}}} +{"q_idx": "detect_000437", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('man wearing red jacket') is boxed.\nIn view 2, locate all target instances 'man wearing green hood' that lie in the direct-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[391.0, 388.0, 458.0, 653.0]], "image_paths": ["images/img_000060.png", "images/img_000069.png"], "supporting": {"draw_boxes": {"images/img_000060.png": [[706, 66, 94, 369]]}}} +{"q_idx": "detect_000438", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('man wearing orange coat') is boxed.\nIn view 3, locate all target instances 'man wearing santa costume' that lie in the direct-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[489.0, 223.0, 526.0, 310.0]], "image_paths": ["images/img_000751.png", "images/img_000739.png", "images/img_000748.png"], "supporting": {"draw_boxes": {"images/img_000739.png": [[334, 140, 70, 163]]}}} +{"q_idx": "detect_000439", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('woman wearing black hat') is boxed.\nIn view 1, locate all target instances 'woman wearing black hat' that lie in the direct-left direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[522.0, 345.0, 546.0, 475.0]], "image_paths": ["images/img_001085.png", "images/img_001090.png", "images/img_001077.png"], "supporting": {"draw_boxes": {"images/img_001077.png": [[725, 19, 75, 448]]}}} +{"q_idx": "detect_000440", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('fan') is boxed.\nIn view 3, locate all target instances 'worker' that lie in the right-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[785.0, 480.0, 815.0, 553.0]], "image_paths": ["images/img_000453.png", "images/img_000461.png", "images/img_000459.png"], "supporting": {"draw_boxes": {"images/img_000453.png": [[497, 215, 62, 62]]}}} +{"q_idx": "detect_000441", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('no hat snowman') is boxed.\nIn view 1, locate all target instances 'red buttons snowman' that lie in the direct-right direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[333.0, 547.0, 381.0, 653.0]], "image_paths": ["images/img_000762.png", "images/img_000756.png"], "supporting": {"draw_boxes": {"images/img_000756.png": [[255, 29, 280, 341]]}}} +{"q_idx": "detect_000442", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('green camouflage helicopter') is boxed.\nIn view 3, locate all target instances 'cream and black striped hot air balloon' that lie in the right-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[205.0, 165.0, 310.0, 383.0]], "image_paths": ["images/img_000981.png", "images/img_000992.png", "images/img_000988.png"], "supporting": {"draw_boxes": {"images/img_000981.png": [[0, 179, 98, 61]]}}} +{"q_idx": "detect_000443", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('stylized F22 fighter jet with camoflague color') is boxed.\nIn view 2, locate all target instances 'green and yellow striped hot air balloon' that lie in the left-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[456.0, 205.0, 541.0, 353.0]], "image_paths": ["images/img_000956.png", "images/img_000958.png", "images/img_000950.png"], "supporting": {"draw_boxes": {"images/img_000950.png": [[426, 136, 235, 112]]}}} +{"q_idx": "detect_000444", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('airplane with red livery') is boxed.\nIn view 2, locate all target instances 'stylized F22 fighter jet with grey and cyan color' that lie in the left-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[691.0, 455.0, 774.0, 515.0]], "image_paths": ["images/img_000320.png", "images/img_000329.png"], "supporting": {"draw_boxes": {"images/img_000320.png": [[0, 109, 128, 103]]}}} +{"q_idx": "detect_000445", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('man in crouching posture') is boxed.\nIn view 1, locate all target instances 'white tombstone' that lie in the direct-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[588.0, 343.0, 626.0, 407.0]], "image_paths": ["images/img_000815.png", "images/img_000806.png"], "supporting": {"draw_boxes": {"images/img_000806.png": [[152, 209, 103, 164]]}}} +{"q_idx": "detect_000446", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('woman in gray clothing') is boxed.\nIn view 2, locate all target instances 'woman in gray clothing' that lie in the left-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[525.0, 437.0, 564.0, 592.0]], "image_paths": ["images/img_001115.png", "images/img_001124.png", "images/img_001121.png"], "supporting": {"draw_boxes": {"images/img_001115.png": [[444, 0, 171, 535]]}}} +{"q_idx": "detect_000447", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('airplane with white and green livery') is boxed.\nIn view 3, locate all target instances 'airplane with yellow livery' that lie in the left-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[359.0, 360.0, 492.0, 420.0], [491.0, 360.0, 625.0, 420.0]], "image_paths": ["images/img_000222.png", "images/img_000231.png", "images/img_000228.png"], "supporting": {"draw_boxes": {"images/img_000222.png": [[19, 161, 279, 77]]}}} +{"q_idx": "detect_000448", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('F-22 fighter jet with gray color') is boxed.\nIn view 3, locate all target instances 'F-22 fighter jet with gray color' that lie in the left-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[695.0, 465.0, 819.0, 583.0]], "image_paths": ["images/img_000905.png", "images/img_000906.png", "images/img_000912.png"], "supporting": {"draw_boxes": {"images/img_000905.png": [[200, 190, 167, 51]]}}} +{"q_idx": "detect_000449", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('white upright tombstone with cross') is boxed.\nIn view 3, locate all target instances 'black flat tombstone with cross' that lie in the right-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[336.0, 563.0, 492.0, 738.0]], "image_paths": ["images/img_001166.png", "images/img_001159.png", "images/img_001168.png"], "supporting": {"draw_boxes": {"images/img_001159.png": [[359, 158, 69, 172]]}}} +{"q_idx": "detect_000450", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('bench made of stone') is boxed.\nIn view 1, locate all target instances 'red curved metal bench without armrests' that lie in the direct-right direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[461.0, 275.0, 546.0, 355.0]], "image_paths": ["images/img_000590.png", "images/img_000581.png"], "supporting": {"draw_boxes": {"images/img_000581.png": [[272, 264, 217, 126]]}}} +{"q_idx": "detect_000451", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('stylized F22 fighter jet with grey and cyan color') is boxed.\nIn view 1, locate all target instances 'F-22 fighter jet with gray color' that lie in the direct-left direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[219.0, 385.0, 329.0, 462.0]], "image_paths": ["images/img_000927.png", "images/img_000929.png", "images/img_000920.png"], "supporting": {"draw_boxes": {"images/img_000920.png": [[0, 156, 146, 86]]}}} +{"q_idx": "detect_000452", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('white upright tombstone with cross') is boxed.\nIn view 3, locate all target instances 'black upright tombstone with cross' that lie in the direct-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[704.0, 443.0, 760.0, 567.0], [690.0, 320.0, 731.0, 445.0]], "image_paths": ["images/img_001170.png", "images/img_001160.png", "images/img_001169.png"], "supporting": {"draw_boxes": {"images/img_001160.png": [[739, 149, 61, 221]]}}} +{"q_idx": "detect_000453", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('man wearing brown coat') is boxed.\nIn view 1, locate all target instances 'man wearing brown coat' that lie in the right-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[522.0, 435.0, 583.0, 668.0]], "image_paths": ["images/img_000043.png", "images/img_000035.png"], "supporting": {"draw_boxes": {"images/img_000035.png": [[331, 68, 108, 333]]}}} +{"q_idx": "detect_000454", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('airplane with yellow livery') is boxed.\nIn view 1, locate all target instances 'airplane with white and green livery' that lie in the left-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[46.0, 453.0, 214.0, 508.0]], "image_paths": ["images/img_000214.png", "images/img_000207.png"], "supporting": {"draw_boxes": {"images/img_000207.png": [[414, 99, 279, 137]]}}} +{"q_idx": "detect_000455", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('person with black hat') is boxed.\nIn view 2, locate all target instances 'person saluting' that lie in the direct-right direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[485.0, 213.0, 512.0, 343.0]], "image_paths": ["images/img_000695.png", "images/img_000704.png", "images/img_000702.png"], "supporting": {"draw_boxes": {"images/img_000695.png": [[91, 40, 126, 367]]}}} +{"q_idx": "detect_000456", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('person saluting') is boxed.\nIn view 1, locate all target instances 'lifeguard tower' that lie in the direct-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[405.0, 2.0, 986.0, 390.0]], "image_paths": ["images/img_000704.png", "images/img_000695.png", "images/img_000697.png"], "supporting": {"draw_boxes": {"images/img_000695.png": [[420, 52, 136, 359]]}}} +{"q_idx": "detect_000457", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('bench made of white painted wood') is boxed.\nIn view 1, locate all target instances 'bench made of white painted wood' that lie in the direct-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[650.0, 390.0, 724.0, 538.0]], "image_paths": ["images/img_000604.png", "images/img_000598.png", "images/img_000602.png"], "supporting": {"draw_boxes": {"images/img_000598.png": [[289, 247, 225, 144]]}}} +{"q_idx": "detect_000458", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('green camouflage helicopter') is boxed.\nIn view 1, locate all target instances 'green camouflage helicopter' that lie in the right-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[228.0, 385.0, 330.0, 452.0]], "image_paths": ["images/img_001043.png", "images/img_001035.png", "images/img_001041.png"], "supporting": {"draw_boxes": {"images/img_001035.png": [[717, 184, 83, 47]]}}} +{"q_idx": "detect_000459", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('bench made of red painted metal') is boxed.\nIn view 1, locate all target instances 'beach ball' that lie in the left-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[557.0, 305.0, 570.0, 322.0], [481.0, 307.0, 494.0, 323.0]], "image_paths": ["images/img_000623.png", "images/img_000615.png", "images/img_000619.png"], "supporting": {"draw_boxes": {"images/img_000615.png": [[642, 289, 158, 91]]}}} +{"q_idx": "detect_000460", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('F-16 fighter jet') is boxed.\nIn view 2, locate all target instances 'airplane with red livery' that lie in the left-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[632.0, 422.0, 772.0, 480.0]], "image_paths": ["images/img_000286.png", "images/img_000287.png", "images/img_000278.png"], "supporting": {"draw_boxes": {"images/img_000278.png": [[520, 202, 112, 31]]}}} +{"q_idx": "detect_000461", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('desert camouflage helicopter') is boxed.\nIn view 2, locate all target instances 'olive green helicopter' that lie in the direct-right direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[72.0, 620.0, 231.0, 688.0]], "image_paths": ["images/img_001018.png", "images/img_001024.png"], "supporting": {"draw_boxes": {"images/img_001018.png": [[455, 175, 186, 68]]}}} +{"q_idx": "detect_000462", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('stylized F22 fighter jet with white and red color') is boxed.\nIn view 3, locate all target instances 'pink and maroon striped hot air balloon' that lie in the left-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[844.0, 305.0, 970.0, 502.0]], "image_paths": ["images/img_000918.png", "images/img_000923.png", "images/img_000928.png"], "supporting": {"draw_boxes": {"images/img_000918.png": [[0, 191, 126, 50]]}}} +{"q_idx": "detect_000463", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('mirage fighter jet') is boxed.\nIn view 3, locate all target instances 'F-16 fighter jet' that lie in the direct-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[356.0, 337.0, 463.0, 407.0]], "image_paths": ["images/img_000895.png", "images/img_000890.png", "images/img_000898.png"], "supporting": {"draw_boxes": {"images/img_000890.png": [[308, 185, 162, 53]]}}} +{"q_idx": "detect_000464", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('stylized F22 fighter jet with camoflague color') is boxed.\nIn view 3, locate all target instances 'airplane with red livery' that lie in the left-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[642.0, 492.0, 785.0, 555.0]], "image_paths": ["images/img_000273.png", "images/img_000265.png", "images/img_000274.png"], "supporting": {"draw_boxes": {"images/img_000265.png": [[707, 210, 93, 25]]}}} +{"q_idx": "detect_000465", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('natural wooden bench with armrests') is boxed.\nIn view 1, locate all target instances 'basketball' that lie in the direct-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[828.0, 615.0, 851.0, 645.0], [835.0, 635.0, 859.0, 667.0]], "image_paths": ["images/img_000604.png", "images/img_000598.png"], "supporting": {"draw_boxes": {"images/img_000598.png": [[0, 219, 178, 171]]}}} +{"q_idx": "detect_000466", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('man wearing black coat') is boxed.\nIn view 2, locate all target instances 'man wearing brown coat' that lie in the direct-left direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[626.0, 352.0, 670.0, 550.0], [654.0, 443.0, 719.0, 648.0]], "image_paths": ["images/img_000010.png", "images/img_000009.png", "images/img_000002.png"], "supporting": {"draw_boxes": {"images/img_000002.png": [[469, 0, 180, 460]]}}} +{"q_idx": "detect_000467", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('desert camouflage helicopter') is boxed.\nIn view 1, locate all target instances 'red and black striped hot air balloon' that lie in the left-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[451.0, 147.0, 544.0, 347.0]], "image_paths": ["images/img_001022.png", "images/img_001014.png", "images/img_001027.png"], "supporting": {"draw_boxes": {"images/img_001014.png": [[701, 179, 99, 62]]}}} +{"q_idx": "detect_000468", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('fan') is boxed.\nIn view 2, locate all target instances 'white stack of sandbags' that lie in the left-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[440.0, 237.0, 565.0, 373.0]], "image_paths": ["images/img_000453.png", "images/img_000460.png"], "supporting": {"draw_boxes": {"images/img_000453.png": [[497, 215, 62, 62]]}}} +{"q_idx": "detect_000469", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('dark grey helicopter') is boxed.\nIn view 3, locate all target instances 'olive green helicopter' that lie in the right-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[230.0, 388.0, 324.0, 452.0]], "image_paths": ["images/img_000985.png", "images/img_000993.png", "images/img_000994.png"], "supporting": {"draw_boxes": {"images/img_000985.png": [[161, 175, 184, 68]]}}} +{"q_idx": "detect_000470", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('black upright tombstone with cross') is boxed.\nIn view 1, locate all target instances 'black upright tombstone with cross' that lie in the direct-left direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[689.0, 307.0, 731.0, 438.0]], "image_paths": ["images/img_001149.png", "images/img_001140.png", "images/img_001152.png"], "supporting": {"draw_boxes": {"images/img_001140.png": [[738, 144, 62, 223]]}}} +{"q_idx": "detect_000471", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('dark grey helicopter') is boxed.\nIn view 1, locate all target instances 'dark grey helicopter' that lie in the left-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[375.0, 347.0, 470.0, 397.0]], "image_paths": ["images/img_001007.png", "images/img_000997.png", "images/img_001004.png"], "supporting": {"draw_boxes": {"images/img_000997.png": [[177, 172, 171, 69]]}}} +{"q_idx": "detect_000472", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('airplane with white and green livery') is boxed.\nIn view 1, locate all target instances 'airplane with yellow livery' that lie in the right-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[644.0, 405.0, 796.0, 473.0], [486.0, 557.0, 704.0, 672.0], [704.0, 557.0, 922.0, 672.0]], "image_paths": ["images/img_000258.png", "images/img_000260.png", "images/img_000252.png"], "supporting": {"draw_boxes": {"images/img_000252.png": [[659, 169, 141, 67]]}}} +{"q_idx": "detect_000473", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('worker') is boxed.\nIn view 2, locate all target instances 'worker' that lie in the direct-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[174.0, 385.0, 198.0, 453.0]], "image_paths": ["images/img_000436.png", "images/img_000443.png", "images/img_000439.png"], "supporting": {"draw_boxes": {"images/img_000436.png": [[339, 196, 35, 81]]}}} +{"q_idx": "detect_000474", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('green camouflage helicopter') is boxed.\nIn view 1, locate all target instances 'desert camouflage helicopter' that lie in the direct-right direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[201.0, 403.0, 314.0, 463.0]], "image_paths": ["images/img_001024.png", "images/img_001018.png", "images/img_001027.png"], "supporting": {"draw_boxes": {"images/img_001018.png": [[716, 184, 84, 47]]}}} +{"q_idx": "detect_000475", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('stylized F22 fighter jet with camoflague color') is boxed.\nIn view 1, locate all target instances 'black and white striped hot air balloon' that lie in the right-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[682.0, 227.0, 776.0, 387.0]], "image_paths": ["images/img_000909.png", "images/img_000901.png", "images/img_000912.png"], "supporting": {"draw_boxes": {"images/img_000901.png": [[0, 194, 90, 52]]}}} +{"q_idx": "detect_000476", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('airplane with light blue livery') is boxed.\nIn view 1, locate all target instances 'F-16 fighter jet' that lie in the left-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[795.0, 577.0, 881.0, 655.0]], "image_paths": ["images/img_000313.png", "images/img_000308.png"], "supporting": {"draw_boxes": {"images/img_000308.png": [[688, 168, 112, 72]]}}} +{"q_idx": "detect_000477", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('woman wearing underwear') is boxed.\nIn view 2, locate all target instances 'woman in pink clothing' that lie in the direct-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[346.0, 403.0, 384.0, 535.0]], "image_paths": ["images/img_001057.png", "images/img_001054.png", "images/img_001048.png"], "supporting": {"draw_boxes": {"images/img_001048.png": [[429, 0, 172, 479]]}}} +{"q_idx": "detect_000478", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('woman in gray clothing') is boxed.\nIn view 2, locate all target instances 'woman wearing underwear' that lie in the direct-left direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[351.0, 373.0, 394.0, 528.0]], "image_paths": ["images/img_001113.png", "images/img_001120.png"], "supporting": {"draw_boxes": {"images/img_001113.png": [[453, 0, 187, 514]]}}} +{"q_idx": "detect_000479", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('airplane with white and orange livery') is boxed.\nIn view 1, locate all target instances 'airplane with yellow livery' that lie in the right-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[735.0, 565.0, 941.0, 682.0], [664.0, 420.0, 807.0, 478.0]], "image_paths": ["images/img_000256.png", "images/img_000249.png"], "supporting": {"draw_boxes": {"images/img_000249.png": [[414, 99, 279, 137]]}}} +{"q_idx": "detect_000480", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('tombstone with shoulder top') is boxed.\nIn view 1, locate all target instances 'man squatting' that lie in the right-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[713.0, 205.0, 755.0, 275.0]], "image_paths": ["images/img_001183.png", "images/img_001174.png"], "supporting": {"draw_boxes": {"images/img_001174.png": [[488, 266, 84, 109]]}}} +{"q_idx": "detect_000481", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('woman in white clothing') is boxed.\nIn view 1, locate all target instances 'woman wearing underwear' that lie in the direct-left direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[561.0, 350.0, 603.0, 498.0]], "image_paths": ["images/img_001090.png", "images/img_001080.png"], "supporting": {"draw_boxes": {"images/img_001080.png": [[309, 65, 130, 337]]}}} +{"q_idx": "detect_000482", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('plaid scarf snowman') is boxed.\nIn view 1, locate all target instances 'red gloves snowman' that lie in the right-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[636.0, 410.0, 724.0, 517.0]], "image_paths": ["images/img_000784.png", "images/img_000779.png", "images/img_000777.png"], "supporting": {"draw_boxes": {"images/img_000777.png": [[0, 125, 136, 251]]}}} +{"q_idx": "detect_000483", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('harrier jumpjet fighter jet') is boxed.\nIn view 3, locate all target instances 'harrier jumpjet fighter jet' that lie in the left-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[578.0, 563.0, 713.0, 672.0]], "image_paths": ["images/img_000937.png", "images/img_000943.png", "images/img_000945.png"], "supporting": {"draw_boxes": {"images/img_000937.png": [[8, 178, 291, 65]]}}} +{"q_idx": "detect_000484", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('black upright tombstone with cross') is boxed.\nIn view 2, locate all target instances 'beige upright tombstone with cross' that lie in the left-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[685.0, 490.0, 744.0, 630.0]], "image_paths": ["images/img_001193.png", "images/img_001200.png", "images/img_001204.png"], "supporting": {"draw_boxes": {"images/img_001193.png": [[1, 158, 55, 213]]}}} +{"q_idx": "detect_000485", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('man wearing green hood') is boxed.\nIn view 2, locate all target instances 'jacket snowman' that lie in the right-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[465.0, 287.0, 507.0, 357.0]], "image_paths": ["images/img_000789.png", "images/img_000796.png", "images/img_000801.png"], "supporting": {"draw_boxes": {"images/img_000789.png": [[72, 120, 103, 201]]}}} +{"q_idx": "detect_000486", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('man in crouching posture') is boxed.\nIn view 1, locate all target instances 'man in sitting posture' that lie in the direct-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[589.0, 438.0, 630.0, 490.0]], "image_paths": ["images/img_000833.png", "images/img_000824.png"], "supporting": {"draw_boxes": {"images/img_000824.png": [[373, 215, 63, 108]]}}} +{"q_idx": "detect_000487", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('lamp') is boxed.\nIn view 2, locate all target instances 'lamp' that lie in the right-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[249.0, 333.0, 276.0, 378.0], [237.0, 370.0, 264.0, 417.0]], "image_paths": ["images/img_000479.png", "images/img_000476.png", "images/img_000466.png"], "supporting": {"draw_boxes": {"images/img_000466.png": [[507, 224, 43, 54]]}}} +{"q_idx": "detect_000488", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('toy crab') is boxed.\nIn view 3, locate all target instances 'gray umbrella' that lie in the left-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[116.0, 37.0, 379.0, 242.0]], "image_paths": ["images/img_000684.png", "images/img_000679.png", "images/img_000686.png"], "supporting": {"draw_boxes": {"images/img_000679.png": [[432, 363, 118, 58]]}}} +{"q_idx": "detect_000489", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('black buttons snowman') is boxed.\nIn view 1, locate all target instances 'black buttons snowman' that lie in the direct-right direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[573.0, 278.0, 674.0, 388.0]], "image_paths": ["images/img_000768.png", "images/img_000763.png", "images/img_000760.png"], "supporting": {"draw_boxes": {"images/img_000760.png": [[239, 38, 320, 334]]}}} +{"q_idx": "detect_000490", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('concrete mixer') is boxed.\nIn view 3, locate all target instances 'concrete mixer' that lie in the right-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[149.0, 717.0, 223.0, 798.0], [212.0, 718.0, 279.0, 800.0]], "image_paths": ["images/img_000436.png", "images/img_000438.png", "images/img_000444.png"], "supporting": {"draw_boxes": {"images/img_000436.png": [[399, 211, 67, 66]]}}} +{"q_idx": "detect_000491", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('beige upright tombstone with cross') is boxed.\nIn view 2, locate all target instances 'colonnade with open top' that lie in the direct-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[0.0, 0.0, 477.0, 467.0]], "image_paths": ["images/img_000870.png", "images/img_000877.png", "images/img_000883.png"], "supporting": {"draw_boxes": {"images/img_000870.png": [[494, 128, 95, 241]]}}} +{"q_idx": "detect_000492", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('person with black hat') is boxed.\nIn view 1, locate all target instances 'toy crab' that lie in the direct-left direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[334.0, 450.0, 389.0, 492.0]], "image_paths": ["images/img_000685.png", "images/img_000679.png", "images/img_000682.png"], "supporting": {"draw_boxes": {"images/img_000679.png": [[710, 72, 90, 335]]}}} +{"q_idx": "detect_000493", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('man wearing black coat') is boxed.\nIn view 3, locate all target instances 'man wearing orange coat' that lie in the direct-right direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[328.0, 250.0, 385.0, 442.0]], "image_paths": ["images/img_000008.png", "images/img_000002.png", "images/img_000007.png"], "supporting": {"draw_boxes": {"images/img_000002.png": [[469, 0, 180, 460]]}}} +{"q_idx": "detect_000494", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('bench made of black metal') is boxed.\nIn view 2, locate all target instances 'baseball' that lie in the right-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[390.0, 313.0, 400.0, 327.0]], "image_paths": ["images/img_000566.png", "images/img_000571.png", "images/img_000564.png"], "supporting": {"draw_boxes": {"images/img_000564.png": [[227, 220, 345, 180]]}}} +{"q_idx": "detect_000495", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('airplane with light blue livery') is boxed.\nIn view 1, locate all target instances 'airplane with red livery' that lie in the right-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[121.0, 468.0, 309.0, 553.0]], "image_paths": ["images/img_000243.png", "images/img_000235.png"], "supporting": {"draw_boxes": {"images/img_000235.png": [[0, 141, 301, 96]]}}} +{"q_idx": "detect_000496", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('F-16 fighter jet') is boxed.\nIn view 1, locate all target instances 'F-16 fighter jet' that lie in the right-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[101.0, 542.0, 245.0, 723.0]], "image_paths": ["images/img_000944.png", "images/img_000939.png", "images/img_000935.png"], "supporting": {"draw_boxes": {"images/img_000935.png": [[93, 172, 193, 67]]}}} +{"q_idx": "detect_000497", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('snowman with black hat and red gloves') is boxed.\nIn view 3, locate all target instances 'snowman with black hat and red gloves' that lie in the direct-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[460.0, 627.0, 549.0, 743.0]], "image_paths": ["images/img_000746.png", "images/img_000743.png", "images/img_000753.png"], "supporting": {"draw_boxes": {"images/img_000743.png": [[280, 98, 219, 239]]}}} +{"q_idx": "detect_000498", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('lamp') is boxed.\nIn view 2, locate all target instances 'truck with white and black cab' that lie in the left-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[138.0, 155.0, 354.0, 368.0]], "image_paths": ["images/img_000438.png", "images/img_000441.png", "images/img_000435.png"], "supporting": {"draw_boxes": {"images/img_000435.png": [[206, 224, 47, 56]]}}} +{"q_idx": "detect_000499", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('natural wooden bench with armrests') is boxed.\nIn view 2, locate all target instances 'bench made of black metal' that lie in the direct-left direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[239.0, 373.0, 333.0, 563.0], [275.0, 272.0, 352.0, 428.0]], "image_paths": ["images/img_000564.png", "images/img_000570.png", "images/img_000571.png"], "supporting": {"draw_boxes": {"images/img_000564.png": [[0, 212, 177, 182]]}}} +{"q_idx": "detect_000500", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('man wearing gray windbreaker') is boxed.\nIn view 3, locate all target instances 'red gloves snowman' that lie in the right-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[636.0, 410.0, 724.0, 517.0], [466.0, 233.0, 535.0, 343.0]], "image_paths": ["images/img_000782.png", "images/img_000774.png", "images/img_000784.png"], "supporting": {"draw_boxes": {"images/img_000774.png": [[722, 125, 78, 229]]}}} +{"q_idx": "detect_000501", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('airplane with red livery') is boxed.\nIn view 3, locate all target instances 'stylized F22 fighter jet with camoflague color' that lie in the left-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[259.0, 525.0, 356.0, 600.0]], "image_paths": ["images/img_000265.png", "images/img_000272.png", "images/img_000270.png"], "supporting": {"draw_boxes": {"images/img_000265.png": [[404, 107, 337, 131]]}}} +{"q_idx": "detect_000502", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('airplane with white and orange livery') is boxed.\nIn view 2, locate all target instances 'airplane with white and orange livery' that lie in the direct-left direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[667.0, 445.0, 807.0, 558.0]], "image_paths": ["images/img_000283.png", "images/img_000284.png", "images/img_000277.png"], "supporting": {"draw_boxes": {"images/img_000277.png": [[123, 154, 319, 84]]}}} +{"q_idx": "detect_000503", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('airplane with red livery') is boxed.\nIn view 2, locate all target instances 'airplane with red livery' that lie in the direct-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[531.0, 388.0, 646.0, 433.0]], "image_paths": ["images/img_000264.png", "images/img_000272.png"], "supporting": {"draw_boxes": {"images/img_000264.png": [[421, 134, 362, 104]]}}} +{"q_idx": "detect_000504", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('airplane with red livery') is boxed.\nIn view 2, locate all target instances 'airplane with light blue livery' that lie in the direct-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[679.0, 430.0, 831.0, 483.0]], "image_paths": ["images/img_000306.png", "images/img_000315.png"], "supporting": {"draw_boxes": {"images/img_000306.png": [[0, 115, 169, 103]]}}} +{"q_idx": "detect_000505", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('airplane with red livery') is boxed.\nIn view 3, locate all target instances 'airplane with red livery' that lie in the right-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[510.0, 565.0, 715.0, 683.0], [664.0, 420.0, 807.0, 478.0]], "image_paths": ["images/img_000245.png", "images/img_000234.png", "images/img_000241.png"], "supporting": {"draw_boxes": {"images/img_000234.png": [[108, 100, 278, 137]]}}} +{"q_idx": "detect_000506", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('bench made of black metal') is boxed.\nIn view 2, locate all target instances 'baseball' that lie in the direct-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[760.0, 365.0, 771.0, 378.0]], "image_paths": ["images/img_000641.png", "images/img_000638.png", "images/img_000628.png"], "supporting": {"draw_boxes": {"images/img_000628.png": [[0, 222, 204, 179]]}}} +{"q_idx": "detect_000507", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('swim ring') is boxed.\nIn view 1, locate all target instances 'surfboard' that lie in the direct-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[689.0, 320.0, 748.0, 543.0]], "image_paths": ["images/img_000668.png", "images/img_000665.png", "images/img_000660.png"], "supporting": {"draw_boxes": {"images/img_000660.png": [[557, 301, 195, 147]]}}} +{"q_idx": "detect_000508", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('striped scarf snowman') is boxed.\nIn view 3, locate all target instances 'jacket snowman' that lie in the direct-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[555.0, 318.0, 604.0, 405.0]], "image_paths": ["images/img_000734.png", "images/img_000722.png", "images/img_000732.png"], "supporting": {"draw_boxes": {"images/img_000722.png": [[396, 135, 141, 202]]}}} +{"q_idx": "detect_000509", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('airplane with red livery') is boxed.\nIn view 2, locate all target instances 'airplane with red livery' that lie in the right-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[206.0, 380.0, 344.0, 487.0], [156.0, 432.0, 318.0, 565.0]], "image_paths": ["images/img_000193.png", "images/img_000202.png"], "supporting": {"draw_boxes": {"images/img_000193.png": [[0, 190, 110, 46]]}}} +{"q_idx": "detect_000510", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('worker') is boxed.\nIn view 2, locate all target instances 'worker' that lie in the direct-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[170.0, 535.0, 201.0, 607.0]], "image_paths": ["images/img_000455.png", "images/img_000459.png", "images/img_000453.png"], "supporting": {"draw_boxes": {"images/img_000453.png": [[253, 196, 37, 82]]}}} +{"q_idx": "detect_000511", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('man wearing green hood') is boxed.\nIn view 2, locate all target instances 'man wearing gray windbreaker' that lie in the right-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[431.0, 260.0, 482.0, 443.0]], "image_paths": ["images/img_000049.png", "images/img_000057.png"], "supporting": {"draw_boxes": {"images/img_000049.png": [[474, 0, 178, 480]]}}} +{"q_idx": "detect_000512", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('man holding box') is boxed.\nIn view 1, locate all target instances 'man holding box' that lie in the right-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[641.0, 432.0, 669.0, 513.0]], "image_paths": ["images/img_000844.png", "images/img_000838.png", "images/img_000850.png"], "supporting": {"draw_boxes": {"images/img_000838.png": [[173, 101, 99, 262]]}}} +{"q_idx": "detect_000513", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('woman in white clothing') is boxed.\nIn view 1, locate all target instances 'woman in white clothing' that lie in the left-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[384.0, 420.0, 436.0, 557.0]], "image_paths": ["images/img_001053.png", "images/img_001045.png", "images/img_001052.png"], "supporting": {"draw_boxes": {"images/img_001045.png": [[149, 0, 159, 472]]}}} +{"q_idx": "detect_000514", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('mirage fighter jet') is boxed.\nIn view 2, locate all target instances 'mirage fighter jet' that lie in the right-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[61.0, 592.0, 281.0, 692.0]], "image_paths": ["images/img_000886.png", "images/img_000895.png", "images/img_000898.png"], "supporting": {"draw_boxes": {"images/img_000886.png": [[476, 118, 199, 124]]}}} +{"q_idx": "detect_000515", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('black tombstone') is boxed.\nIn view 2, locate all target instances 'tombstone with pointed top' that lie in the direct-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[215.0, 630.0, 274.0, 768.0], [362.0, 638.0, 441.0, 750.0]], "image_paths": ["images/img_001172.png", "images/img_001182.png"], "supporting": {"draw_boxes": {"images/img_001172.png": [[653, 302, 117, 53]]}}} +{"q_idx": "detect_000516", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('F-16 fighter jet') is boxed.\nIn view 2, locate all target instances 'pink and maroon striped hot air balloon' that lie in the right-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[848.0, 305.0, 975.0, 505.0]], "image_paths": ["images/img_000948.png", "images/img_000958.png"], "supporting": {"draw_boxes": {"images/img_000948.png": [[0, 203, 85, 35]]}}} +{"q_idx": "detect_000517", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('airplane with yellow livery') is boxed.\nIn view 1, locate all target instances 'airplane with white and green livery' that lie in the left-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[60.0, 565.0, 266.0, 682.0]], "image_paths": ["images/img_000213.png", "images/img_000207.png", "images/img_000218.png"], "supporting": {"draw_boxes": {"images/img_000207.png": [[414, 99, 279, 137]]}}} +{"q_idx": "detect_000518", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('woman wearing black dress') is boxed.\nIn view 2, locate all target instances 'woman in pink clothing' that lie in the left-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[449.0, 450.0, 494.0, 590.0]], "image_paths": ["images/img_001068.png", "images/img_001067.png", "images/img_001059.png"], "supporting": {"draw_boxes": {"images/img_001059.png": [[0, 44, 88, 421]]}}} +{"q_idx": "detect_000519", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('man wearing black coat') is boxed.\nIn view 1, locate all target instances 'weathered metal bus shelter with side walls' that lie in the direct-right direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[591.0, 0.0, 1000.0, 503.0]], "image_paths": ["images/img_000014.png", "images/img_000004.png"], "supporting": {"draw_boxes": {"images/img_000004.png": [[0, 0, 167, 559]]}}} +{"q_idx": "detect_000520", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('stylized F22 fighter jet with grey and cyan color') is boxed.\nIn view 2, locate all target instances 'stylized F22 fighter jet with grey and cyan color' that lie in the right-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[254.0, 328.0, 356.0, 415.0], [384.0, 328.0, 484.0, 415.0]], "image_paths": ["images/img_000950.png", "images/img_000960.png", "images/img_000955.png"], "supporting": {"draw_boxes": {"images/img_000950.png": [[113, 139, 273, 109]]}}} +{"q_idx": "detect_000521", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('mirage fighter jet') is boxed.\nIn view 1, locate all target instances 'airplane with white and orange livery' that lie in the right-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[625.0, 423.0, 764.0, 480.0], [186.0, 482.0, 350.0, 560.0]], "image_paths": ["images/img_000301.png", "images/img_000295.png"], "supporting": {"draw_boxes": {"images/img_000295.png": [[170, 199, 114, 35]]}}} +{"q_idx": "detect_000522", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('bench made of stone') is boxed.\nIn view 2, locate all target instances 'tennis ball' that lie in the right-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[690.0, 737.0, 705.0, 757.0]], "image_paths": ["images/img_000581.png", "images/img_000588.png"], "supporting": {"draw_boxes": {"images/img_000581.png": [[272, 264, 217, 126]]}}} +{"q_idx": "detect_000523", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('olive green helicopter') is boxed.\nIn view 1, locate all target instances 'pink and maroon striped hot air balloon' that lie in the right-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[203.0, 205.0, 307.0, 382.0]], "image_paths": ["images/img_000992.png", "images/img_000990.png", "images/img_000985.png"], "supporting": {"draw_boxes": {"images/img_000985.png": [[504, 178, 78, 65]]}}} +{"q_idx": "detect_000524", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('airplane with red livery') is boxed.\nIn view 1, locate all target instances 'airplane with yellow livery' that lie in the direct-right direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[667.0, 468.0, 846.0, 553.0]], "image_paths": ["images/img_000243.png", "images/img_000234.png"], "supporting": {"draw_boxes": {"images/img_000234.png": [[108, 100, 278, 137]]}}} +{"q_idx": "detect_000525", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('person saluting') is boxed.\nIn view 2, locate all target instances 'person saluting' that lie in the direct-right direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[340.0, 243.0, 370.0, 375.0]], "image_paths": ["images/img_000699.png", "images/img_000703.png", "images/img_000695.png"], "supporting": {"draw_boxes": {"images/img_000695.png": [[420, 52, 136, 359]]}}} +{"q_idx": "detect_000526", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('man wearing red jacket') is boxed.\nIn view 2, locate all target instances 'bus stop without enclosure' that lie in the left-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[564.0, 48.0, 974.0, 395.0]], "image_paths": ["images/img_000060.png", "images/img_000068.png", "images/img_000065.png"], "supporting": {"draw_boxes": {"images/img_000060.png": [[706, 66, 94, 369]]}}} +{"q_idx": "detect_000527", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('man wearing a green hoodie and orange shoes') is boxed.\nIn view 1, locate all target instances 'colonnade with circular top' that lie in the right-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[155.0, 0.0, 821.0, 317.0]], "image_paths": ["images/img_000878.png", "images/img_000871.png"], "supporting": {"draw_boxes": {"images/img_000871.png": [[346, 111, 82, 220]]}}} +{"q_idx": "detect_000528", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('natural wooden bench with armrests') is boxed.\nIn view 2, locate all target instances 'natural wooden bench with armrests' that lie in the right-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[205.0, 548.0, 301.0, 745.0]], "image_paths": ["images/img_000607.png", "images/img_000604.png", "images/img_000594.png"], "supporting": {"draw_boxes": {"images/img_000594.png": [[0, 229, 172, 164]]}}} +{"q_idx": "detect_000529", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('woman wearing scarf accessory') is boxed.\nIn view 1, locate all target instances 'vendor cart with red and yellow roof' that lie in the left-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[646.0, 247.0, 746.0, 417.0]], "image_paths": ["images/img_001107.png", "images/img_001098.png"], "supporting": {"draw_boxes": {"images/img_001098.png": [[0, 24, 69, 460]]}}} +{"q_idx": "detect_000530", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('man wearing a green hoodie and orange shoes') is boxed.\nIn view 1, locate all target instances 'man wearing a green hoodie and orange shoes' that lie in the right-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[375.0, 445.0, 401.0, 545.0]], "image_paths": ["images/img_000864.png", "images/img_000859.png", "images/img_000858.png"], "supporting": {"draw_boxes": {"images/img_000858.png": [[565, 64, 142, 309]]}}} +{"q_idx": "detect_000531", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('airplane with light blue livery') is boxed.\nIn view 2, locate all target instances 'airplane with red livery' that lie in the direct-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[46.0, 453.0, 214.0, 508.0]], "image_paths": ["images/img_000193.png", "images/img_000199.png"], "supporting": {"draw_boxes": {"images/img_000193.png": [[637, 147, 163, 88]]}}} +{"q_idx": "detect_000532", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('green camouflage helicopter') is boxed.\nIn view 3, locate all target instances 'green camouflage helicopter' that lie in the direct-right direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[458.0, 312.0, 537.0, 362.0]], "image_paths": ["images/img_000980.png", "images/img_000990.png", "images/img_000988.png"], "supporting": {"draw_boxes": {"images/img_000980.png": [[86, 170, 196, 71]]}}} +{"q_idx": "detect_000533", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('person with tattoos') is boxed.\nIn view 1, locate all target instances 'toy crab' that lie in the right-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[427.0, 333.0, 472.0, 373.0], [307.0, 443.0, 359.0, 490.0]], "image_paths": ["images/img_000720.png", "images/img_000712.png", "images/img_000713.png"], "supporting": {"draw_boxes": {"images/img_000712.png": [[448, 75, 102, 310]]}}} +{"q_idx": "detect_000534", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('stylized F22 fighter jet with camoflague color') is boxed.\nIn view 3, locate all target instances 'cream and black striped hot air balloon' that lie in the left-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[19.0, 257.0, 147.0, 500.0]], "image_paths": ["images/img_000910.png", "images/img_000903.png", "images/img_000908.png"], "supporting": {"draw_boxes": {"images/img_000903.png": [[658, 154, 141, 87]]}}} +{"q_idx": "detect_000535", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('olive green helicopter') is boxed.\nIn view 2, locate all target instances 'olive green helicopter' that lie in the left-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[121.0, 587.0, 186.0, 712.0]], "image_paths": ["images/img_000972.png", "images/img_000973.png", "images/img_000965.png"], "supporting": {"draw_boxes": {"images/img_000965.png": [[214, 171, 82, 70]]}}} +{"q_idx": "detect_000536", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('toy crab') is boxed.\nIn view 1, locate all target instances 'toy crab' that lie in the direct-left direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[421.0, 650.0, 477.0, 713.0]], "image_paths": ["images/img_000718.png", "images/img_000712.png", "images/img_000716.png"], "supporting": {"draw_boxes": {"images/img_000712.png": [[244, 347, 117, 40]]}}} +{"q_idx": "detect_000537", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('woman wearing black hat') is boxed.\nIn view 1, locate all target instances 'woman wearing black hat' that lie in the right-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[477.0, 492.0, 511.0, 643.0]], "image_paths": ["images/img_001088.png", "images/img_001081.png"], "supporting": {"draw_boxes": {"images/img_001081.png": [[172, 0, 139, 486]]}}} +{"q_idx": "detect_000538", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('woman in gray clothing') is boxed.\nIn view 1, locate all target instances 'woman wearing underwear' that lie in the direct-left direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[470.0, 313.0, 495.0, 452.0]], "image_paths": ["images/img_001122.png", "images/img_001113.png", "images/img_001119.png"], "supporting": {"draw_boxes": {"images/img_001113.png": [[453, 0, 187, 514]]}}} +{"q_idx": "detect_000539", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('tombstone with round top') is boxed.\nIn view 2, locate all target instances 'beige upright tombstone with cross' that lie in the direct-left direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[751.0, 385.0, 812.0, 523.0]], "image_paths": ["images/img_001192.png", "images/img_001202.png"], "supporting": {"draw_boxes": {"images/img_001192.png": [[541, 258, 171, 139]]}}} +{"q_idx": "detect_000540", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('man wearing santa costume') is boxed.\nIn view 1, locate all target instances 'man wearing santa costume' that lie in the direct-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[264.0, 327.0, 336.0, 542.0]], "image_paths": ["images/img_000053.png", "images/img_000047.png"], "supporting": {"draw_boxes": {"images/img_000047.png": [[457, 0, 229, 472]]}}} +{"q_idx": "detect_000541", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('worker') is boxed.\nIn view 2, locate all target instances 'truck with blue and white cab' that lie in the left-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[0.0, 203.0, 175.0, 537.0]], "image_paths": ["images/img_000437.png", "images/img_000444.png"], "supporting": {"draw_boxes": {"images/img_000437.png": [[652, 190, 54, 97]]}}} +{"q_idx": "detect_000542", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('white tombstone') is boxed.\nIn view 2, locate all target instances 'man wearing a dark kurta and white pants' that lie in the left-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[369.0, 452.0, 394.0, 540.0]], "image_paths": ["images/img_000837.png", "images/img_000845.png"], "supporting": {"draw_boxes": {"images/img_000837.png": [[493, 128, 87, 238]]}}} +{"q_idx": "detect_000543", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('green camouflage helicopter') is boxed.\nIn view 1, locate all target instances 'dark grey helicopter' that lie in the direct-right direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[75.0, 433.0, 184.0, 512.0], [357.0, 345.0, 445.0, 405.0]], "image_paths": ["images/img_001041.png", "images/img_001036.png", "images/img_001035.png"], "supporting": {"draw_boxes": {"images/img_001035.png": [[717, 184, 83, 47]]}}} +{"q_idx": "detect_000544", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('black tombstone') is boxed.\nIn view 1, locate all target instances 'white tombstone' that lie in the direct-right direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[432.0, 360.0, 509.0, 420.0]], "image_paths": ["images/img_000859.png", "images/img_000852.png"], "supporting": {"draw_boxes": {"images/img_000852.png": [[269, 232, 206, 125]]}}} +{"q_idx": "detect_000545", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('woman in pink clothing') is boxed.\nIn view 2, locate all target instances 'woman in pink clothing' that lie in the direct-left direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[458.0, 362.0, 487.0, 482.0]], "image_paths": ["images/img_001069.png", "images/img_001070.png", "images/img_001060.png"], "supporting": {"draw_boxes": {"images/img_001060.png": [[480, 0, 271, 541]]}}} +{"q_idx": "detect_000546", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('man wearing a bandana and a denim vest, sitting') is boxed.\nIn view 3, locate all target instances 'man in a plaid shirt with his arms crossed' that lie in the direct-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[334.0, 537.0, 366.0, 627.0]], "image_paths": ["images/img_000874.png", "images/img_000882.png", "images/img_000880.png"], "supporting": {"draw_boxes": {"images/img_000874.png": [[207, 236, 140, 135]]}}} +{"q_idx": "detect_000547", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('stylized F22 fighter jet with grey and cyan color') is boxed.\nIn view 3, locate all target instances 'stylized F22 fighter jet with camoflague color' that lie in the direct-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[354.0, 332.0, 468.0, 417.0]], "image_paths": ["images/img_000924.png", "images/img_000921.png", "images/img_000929.png"], "supporting": {"draw_boxes": {"images/img_000921.png": [[318, 173, 165, 67]]}}} +{"q_idx": "detect_000548", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('man wearing orange coat') is boxed.\nIn view 2, locate all target instances 'red gloves snowman' that lie in the right-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[449.0, 620.0, 535.0, 737.0], [566.0, 337.0, 645.0, 437.0], [634.0, 412.0, 719.0, 517.0]], "image_paths": ["images/img_000778.png", "images/img_000786.png"], "supporting": {"draw_boxes": {"images/img_000778.png": [[626, 102, 95, 239]]}}} +{"q_idx": "detect_000549", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('F-16 fighter jet') is boxed.\nIn view 3, locate all target instances 'airplane with white and orange livery' that lie in the left-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[532.0, 390.0, 647.0, 433.0]], "image_paths": ["images/img_000291.png", "images/img_000304.png", "images/img_000300.png"], "supporting": {"draw_boxes": {"images/img_000291.png": [[107, 201, 135, 33]]}}} +{"q_idx": "detect_000550", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('woman in gray clothing') is boxed.\nIn view 1, locate all target instances 'woman in pink clothing' that lie in the direct-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[475.0, 323.0, 532.0, 447.0]], "image_paths": ["images/img_001056.png", "images/img_001050.png"], "supporting": {"draw_boxes": {"images/img_001050.png": [[434, 0, 208, 475]]}}} +{"q_idx": "detect_000551", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('man holding box') is boxed.\nIn view 2, locate all target instances 'man wearing a green hoodie and orange shoes' that lie in the left-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[346.0, 437.0, 376.0, 533.0]], "image_paths": ["images/img_000853.png", "images/img_000863.png"], "supporting": {"draw_boxes": {"images/img_000853.png": [[0, 111, 79, 247]]}}} +{"q_idx": "detect_000552", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('toy crab') is boxed.\nIn view 1, locate all target instances 'blue swim ring' that lie in the direct-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[253.0, 477.0, 311.0, 537.0]], "image_paths": ["images/img_000686.png", "images/img_000679.png", "images/img_000688.png"], "supporting": {"draw_boxes": {"images/img_000679.png": [[432, 363, 118, 58]]}}} +{"q_idx": "detect_000553", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('man wearing brown coat') is boxed.\nIn view 3, locate all target instances 'man wearing red jacket' that lie in the direct-right direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[380.0, 287.0, 430.0, 478.0]], "image_paths": ["images/img_000038.png", "images/img_000032.png", "images/img_000039.png"], "supporting": {"draw_boxes": {"images/img_000032.png": [[541, 0, 145, 458]]}}} +{"q_idx": "detect_000554", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('white flat tombstone with cross') is boxed.\nIn view 2, locate all target instances 'black flat tombstone with cross' that lie in the right-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[265.0, 367.0, 400.0, 505.0]], "image_paths": ["images/img_001156.png", "images/img_001164.png"], "supporting": {"draw_boxes": {"images/img_001156.png": [[449, 218, 265, 213]]}}} +{"q_idx": "detect_000555", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('tombstone with pointed top') is boxed.\nIn view 2, locate all target instances 'tombstone with pointed top' that lie in the left-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[495.0, 748.0, 581.0, 905.0]], "image_paths": ["images/img_001130.png", "images/img_001137.png"], "supporting": {"draw_boxes": {"images/img_001130.png": [[709, 191, 89, 168]]}}} +{"q_idx": "detect_000556", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('no hat snowman') is boxed.\nIn view 3, locate all target instances 'no hat snowman' that lie in the left-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[559.0, 557.0, 651.0, 668.0]], "image_paths": ["images/img_000757.png", "images/img_000762.png", "images/img_000764.png"], "supporting": {"draw_boxes": {"images/img_000757.png": [[567, 0, 233, 410]]}}} +{"q_idx": "detect_000557", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('beige sand helicopter') is boxed.\nIn view 2, locate all target instances 'beige sand helicopter' that lie in the direct-left direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[662.0, 472.0, 865.0, 557.0], [621.0, 340.0, 767.0, 398.0]], "image_paths": ["images/img_000973.png", "images/img_000975.png", "images/img_000965.png"], "supporting": {"draw_boxes": {"images/img_000965.png": [[657, 179, 143, 62]]}}} +{"q_idx": "detect_000558", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('olive green helicopter') is boxed.\nIn view 2, locate all target instances 'dark grey helicopter' that lie in the right-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[165.0, 465.0, 278.0, 562.0], [542.0, 565.0, 680.0, 693.0]], "image_paths": ["images/img_001000.png", "images/img_001009.png", "images/img_001010.png"], "supporting": {"draw_boxes": {"images/img_001000.png": [[373, 198, 96, 41]]}}} +{"q_idx": "detect_000559", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('worker') is boxed.\nIn view 3, locate all target instances 'worker' that lie in the direct-right direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[785.0, 480.0, 815.0, 553.0]], "image_paths": ["images/img_000457.png", "images/img_000453.png", "images/img_000459.png"], "supporting": {"draw_boxes": {"images/img_000453.png": [[253, 196, 37, 82]]}}} +{"q_idx": "detect_000560", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('worker') is boxed.\nIn view 2, locate all target instances 'concrete mixer' that lie in the left-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[696.0, 593.0, 743.0, 672.0]], "image_paths": ["images/img_000442.png", "images/img_000445.png", "images/img_000436.png"], "supporting": {"draw_boxes": {"images/img_000436.png": [[339, 196, 35, 81]]}}} +{"q_idx": "detect_000561", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('white upright tombstone with cross') is boxed.\nIn view 2, locate all target instances 'tombstone with pointed top' that lie in the left-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[549.0, 255.0, 608.0, 353.0]], "image_paths": ["images/img_001176.png", "images/img_001186.png", "images/img_001187.png"], "supporting": {"draw_boxes": {"images/img_001176.png": [[495, 128, 89, 238]]}}} +{"q_idx": "detect_000562", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('mirage fighter jet') is boxed.\nIn view 1, locate all target instances 'mirage fighter jet' that lie in the right-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[539.0, 368.0, 639.0, 417.0]], "image_paths": ["images/img_000925.png", "images/img_000929.png", "images/img_000919.png"], "supporting": {"draw_boxes": {"images/img_000919.png": [[484, 166, 315, 76]]}}} +{"q_idx": "detect_000563", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('tombstone with pointed top') is boxed.\nIn view 1, locate all target instances 'tombstone with pointed top' that lie in the right-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[429.0, 710.0, 511.0, 860.0]], "image_paths": ["images/img_001133.png", "images/img_001138.png", "images/img_001125.png"], "supporting": {"draw_boxes": {"images/img_001125.png": [[0, 192, 90, 168]]}}} +{"q_idx": "detect_000564", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('bench made of black metal') is boxed.\nIn view 2, locate all target instances 'baseball' that lie in the left-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[160.0, 603.0, 174.0, 622.0]], "image_paths": ["images/img_000623.png", "images/img_000621.png", "images/img_000611.png"], "supporting": {"draw_boxes": {"images/img_000611.png": [[225, 222, 346, 186]]}}} +{"q_idx": "detect_000565", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('worker') is boxed.\nIn view 2, locate all target instances 'truck with white and black cab' that lie in the left-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[150.0, 147.0, 359.0, 352.0]], "image_paths": ["images/img_000453.png", "images/img_000460.png", "images/img_000461.png"], "supporting": {"draw_boxes": {"images/img_000453.png": [[253, 196, 37, 82]]}}} +{"q_idx": "detect_000566", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('airplane with white and green livery') is boxed.\nIn view 2, locate all target instances 'airplane with yellow livery' that lie in the direct-left direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[506.0, 373.0, 630.0, 418.0]], "image_paths": ["images/img_000218.png", "images/img_000213.png", "images/img_000206.png"], "supporting": {"draw_boxes": {"images/img_000206.png": [[349, 159, 166, 75]]}}} +{"q_idx": "detect_000567", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('white painted wooden bench with armrests') is boxed.\nIn view 1, locate all target instances 'basketball' that lie in the left-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[736.0, 758.0, 762.0, 793.0]], "image_paths": ["images/img_000606.png", "images/img_000605.png", "images/img_000598.png"], "supporting": {"draw_boxes": {"images/img_000598.png": [[289, 247, 225, 144]]}}} +{"q_idx": "detect_000568", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('woman wearing scarf accessory') is boxed.\nIn view 2, locate all target instances 'woman wearing scarf accessory' that lie in the direct-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[480.0, 333.0, 501.0, 453.0]], "image_paths": ["images/img_001098.png", "images/img_001107.png"], "supporting": {"draw_boxes": {"images/img_001098.png": [[721, 36, 79, 434]]}}} +{"q_idx": "detect_000569", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('mirage fighter jet') is boxed.\nIn view 2, locate all target instances 'F-16 fighter jet' that lie in the direct-left direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[554.0, 337.0, 647.0, 407.0]], "image_paths": ["images/img_000890.png", "images/img_000898.png"], "supporting": {"draw_boxes": {"images/img_000890.png": [[308, 185, 162, 53]]}}} +{"q_idx": "detect_000570", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('lamp') is boxed.\nIn view 3, locate all target instances 'forklift' that lie in the right-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[196.0, 215.0, 302.0, 372.0]], "image_paths": ["images/img_000465.png", "images/img_000479.png", "images/img_000473.png"], "supporting": {"draw_boxes": {"images/img_000465.png": [[217, 224, 46, 56]]}}} +{"q_idx": "detect_000571", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('bench made of stone') is boxed.\nIn view 3, locate all target instances 'soccer ball' that lie in the right-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[836.0, 675.0, 860.0, 705.0]], "image_paths": ["images/img_000586.png", "images/img_000581.png", "images/img_000587.png"], "supporting": {"draw_boxes": {"images/img_000581.png": [[272, 264, 217, 126]]}}} +{"q_idx": "detect_000572", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('airplane with yellow livery') is boxed.\nIn view 1, locate all target instances 'airplane with white and green livery' that lie in the left-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[331.0, 378.0, 449.0, 423.0]], "image_paths": ["images/img_000257.png", "images/img_000256.png", "images/img_000251.png"], "supporting": {"draw_boxes": {"images/img_000251.png": [[19, 161, 279, 77]]}}} +{"q_idx": "detect_000573", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('man wearing black coat') is boxed.\nIn view 3, locate all target instances 'man wearing black coat' that lie in the right-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[465.0, 423.0, 535.0, 673.0]], "image_paths": ["images/img_000023.png", "images/img_000020.png", "images/img_000027.png"], "supporting": {"draw_boxes": {"images/img_000020.png": [[523, 0, 153, 451]]}}} +{"q_idx": "detect_000574", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('tombstone without cross') is boxed.\nIn view 1, locate all target instances 'beige tombstone' that lie in the direct-left direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[504.0, 155.0, 545.0, 283.0]], "image_paths": ["images/img_001150.png", "images/img_001144.png"], "supporting": {"draw_boxes": {"images/img_001144.png": [[461, 190, 146, 179]]}}} +{"q_idx": "detect_000575", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('worker wearing red hat and white shirt') is boxed.\nIn view 2, locate all target instances 'worker wearing red hat and white shirt' that lie in the left-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[491.0, 288.0, 505.0, 353.0]], "image_paths": ["images/img_000466.png", "images/img_000472.png", "images/img_000473.png"], "supporting": {"draw_boxes": {"images/img_000466.png": [[250, 193, 34, 84]]}}} +{"q_idx": "detect_000576", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('lamp') is boxed.\nIn view 1, locate all target instances 'lamp' that lie in the left-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[249.0, 333.0, 276.0, 378.0], [237.0, 370.0, 264.0, 417.0]], "image_paths": ["images/img_000476.png", "images/img_000466.png", "images/img_000477.png"], "supporting": {"draw_boxes": {"images/img_000466.png": [[671, 224, 60, 53]]}}} +{"q_idx": "detect_000577", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('red curved metal bench without armrests') is boxed.\nIn view 2, locate all target instances 'beach ball' that lie in the left-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[728.0, 380.0, 741.0, 398.0], [660.0, 345.0, 674.0, 362.0]], "image_paths": ["images/img_000615.png", "images/img_000624.png"], "supporting": {"draw_boxes": {"images/img_000615.png": [[642, 289, 158, 91]]}}} +{"q_idx": "detect_000578", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('tombstone with shoulder top') is boxed.\nIn view 1, locate all target instances 'beige upright tombstone with cross' that lie in the direct-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[203.0, 395.0, 260.0, 532.0]], "image_paths": ["images/img_001198.png", "images/img_001199.png", "images/img_001191.png"], "supporting": {"draw_boxes": {"images/img_001191.png": [[721, 264, 78, 107]]}}} +{"q_idx": "detect_000579", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('man wearing gray hat') is boxed.\nIn view 2, locate all target instances 'man wearing gray hat' that lie in the direct-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[259.0, 387.0, 304.0, 487.0]], "image_paths": ["images/img_000727.png", "images/img_000733.png", "images/img_000731.png"], "supporting": {"draw_boxes": {"images/img_000727.png": [[241, 142, 60, 182]]}}} +{"q_idx": "detect_000580", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('harrier jumpjet fighter jet') is boxed.\nIn view 1, locate all target instances 'F-16 fighter jet' that lie in the direct-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[463.0, 330.0, 555.0, 373.0]], "image_paths": ["images/img_000939.png", "images/img_000940.png", "images/img_000930.png"], "supporting": {"draw_boxes": {"images/img_000930.png": [[412, 180, 253, 63]]}}} +{"q_idx": "detect_000581", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('beige sand helicopter') is boxed.\nIn view 1, locate all target instances 'desert camouflage helicopter' that lie in the left-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[721.0, 482.0, 855.0, 557.0]], "image_paths": ["images/img_001020.png", "images/img_001012.png", "images/img_001023.png"], "supporting": {"draw_boxes": {"images/img_001012.png": [[417, 168, 275, 74]]}}} +{"q_idx": "detect_000582", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('lamp') is boxed.\nIn view 1, locate all target instances 'lamp' that lie in the direct-right direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[264.0, 352.0, 294.0, 400.0]], "image_paths": ["images/img_000473.png", "images/img_000465.png", "images/img_000479.png"], "supporting": {"draw_boxes": {"images/img_000465.png": [[217, 224, 46, 56]]}}} +{"q_idx": "detect_000583", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('F-16 fighter jet') is boxed.\nIn view 2, locate all target instances 'F-22 fighter jet with gray color' that lie in the right-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[661.0, 415.0, 804.0, 477.0]], "image_paths": ["images/img_000948.png", "images/img_000955.png", "images/img_000954.png"], "supporting": {"draw_boxes": {"images/img_000948.png": [[0, 203, 85, 35]]}}} +{"q_idx": "detect_000584", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('green camouflage helicopter') is boxed.\nIn view 1, locate all target instances 'green camouflage helicopter' that lie in the direct-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[729.0, 470.0, 846.0, 570.0]], "image_paths": ["images/img_001042.png", "images/img_001043.png", "images/img_001034.png"], "supporting": {"draw_boxes": {"images/img_001034.png": [[352, 192, 120, 46]]}}} +{"q_idx": "detect_000585", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('woman wearing black hat') is boxed.\nIn view 2, locate all target instances 'vendor cart with red and yellow roof' that lie in the left-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[450.0, 212.0, 541.0, 362.0]], "image_paths": ["images/img_001098.png", "images/img_001106.png"], "supporting": {"draw_boxes": {"images/img_001098.png": [[0, 24, 69, 460]]}}} +{"q_idx": "detect_000586", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('red gloves snowman') is boxed.\nIn view 2, locate all target instances 'red gloves snowman' that lie in the direct-right direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[466.0, 233.0, 535.0, 343.0]], "image_paths": ["images/img_000776.png", "images/img_000784.png", "images/img_000779.png"], "supporting": {"draw_boxes": {"images/img_000776.png": [[0, 0, 273, 411]]}}} +{"q_idx": "detect_000587", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('man wearing green hood') is boxed.\nIn view 3, locate all target instances 'man wearing santa costume' that lie in the direct-right direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[526.0, 423.0, 613.0, 663.0]], "image_paths": ["images/img_000068.png", "images/img_000062.png", "images/img_000070.png"], "supporting": {"draw_boxes": {"images/img_000062.png": [[704, 5, 96, 447]]}}} +{"q_idx": "detect_000588", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('stylized F22 fighter jet with camoflague color') is boxed.\nIn view 1, locate all target instances 'airplane with red livery' that lie in the direct-left direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[325.0, 550.0, 485.0, 632.0]], "image_paths": ["images/img_000272.png", "images/img_000264.png"], "supporting": {"draw_boxes": {"images/img_000264.png": [[405, 194, 102, 42]]}}} +{"q_idx": "detect_000589", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('worker') is boxed.\nIn view 3, locate all target instances 'truck with white and black cab' that lie in the left-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[465.0, 92.0, 576.0, 290.0]], "image_paths": ["images/img_000455.png", "images/img_000453.png", "images/img_000461.png"], "supporting": {"draw_boxes": {"images/img_000453.png": [[253, 196, 37, 82]]}}} +{"q_idx": "detect_000590", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('black upright tombstone with cross') is boxed.\nIn view 2, locate all target instances 'black upright tombstone with cross' that lie in the direct-left direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[362.0, 607.0, 421.0, 755.0]], "image_paths": ["images/img_001145.png", "images/img_001152.png"], "supporting": {"draw_boxes": {"images/img_001145.png": [[0, 151, 66, 215]]}}} +{"q_idx": "detect_000591", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('white upright tombstone with cross') is boxed.\nIn view 2, locate all target instances 'gray colonnade with circular top' that lie in the left-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[246.0, 0.0, 756.0, 295.0]], "image_paths": ["images/img_000821.png", "images/img_000829.png", "images/img_000827.png"], "supporting": {"draw_boxes": {"images/img_000821.png": [[748, 157, 51, 216]]}}} +{"q_idx": "detect_000592", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('man holding box') is boxed.\nIn view 1, locate all target instances 'colonnade with rectangular top' that lie in the right-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[339.0, 48.0, 660.0, 320.0]], "image_paths": ["images/img_000847.png", "images/img_000839.png", "images/img_000850.png"], "supporting": {"draw_boxes": {"images/img_000839.png": [[492, 105, 87, 259]]}}} +{"q_idx": "detect_000593", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('green camouflage helicopter') is boxed.\nIn view 2, locate all target instances 'dark grey helicopter' that lie in the direct-right direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[168.0, 482.0, 296.0, 555.0]], "image_paths": ["images/img_000997.png", "images/img_001003.png"], "supporting": {"draw_boxes": {"images/img_000997.png": [[0, 179, 98, 61]]}}} +{"q_idx": "detect_000594", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('man wearing brown coat') is boxed.\nIn view 1, locate all target instances 'man wearing brown coat' that lie in the direct-behind direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[260.0, 442.0, 333.0, 672.0]], "image_paths": ["images/img_000042.png", "images/img_000035.png", "images/img_000037.png"], "supporting": {"draw_boxes": {"images/img_000035.png": [[331, 68, 108, 333]]}}} +{"q_idx": "detect_000595", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('surfboard') is boxed.\nIn view 2, locate all target instances 'surfboard' that lie in the left-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[594.0, 182.0, 655.0, 403.0]], "image_paths": ["images/img_000668.png", "images/img_000667.png", "images/img_000660.png"], "supporting": {"draw_boxes": {"images/img_000660.png": [[227, 0, 155, 396]]}}} +{"q_idx": "detect_000596", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('man wearing gray hat') is boxed.\nIn view 1, locate all target instances 'man wearing red jacket' that lie in the direct-left direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[510.0, 253.0, 562.0, 448.0]], "image_paths": ["images/img_000043.png", "images/img_000034.png"], "supporting": {"draw_boxes": {"images/img_000034.png": [[666, 45, 134, 396]]}}} +{"q_idx": "detect_000597", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('airplane with white and green livery') is boxed.\nIn view 2, locate all target instances 'airplane with white and green livery' that lie in the direct-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[205.0, 380.0, 343.0, 487.0]], "image_paths": ["images/img_000223.png", "images/img_000230.png", "images/img_000231.png"], "supporting": {"draw_boxes": {"images/img_000223.png": [[0, 168, 141, 67]]}}} +{"q_idx": "detect_000598", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find the reference instance ('surfboard') is boxed.\nIn view 2, locate all target instances 'surfboard' that lie in the right-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[619.0, 415.0, 681.0, 652.0]], "image_paths": ["images/img_000670.png", "images/img_000671.png", "images/img_000663.png"], "supporting": {"draw_boxes": {"images/img_000663.png": [[618, 0, 156, 399]]}}} +{"q_idx": "detect_000599", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find the reference instance ('plaid scarf snowman') is boxed.\nIn view 2, locate all target instances 'red buttons snowman' that lie in the right-front direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[482.0, 623.0, 539.0, 715.0]], "image_paths": ["images/img_000789.png", "images/img_000796.png"], "supporting": {"draw_boxes": {"images/img_000789.png": [[334, 133, 137, 236]]}}} +{"q_idx": "detect_000600", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find the reference instance ('black upright tombstone with cross') is boxed.\nIn view 1, locate all target instances 'black flat tombstone with cross' that lie in the direct-left direction relative to the reference instance.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[522.0, 583.0, 601.0, 757.0]], "image_paths": ["images/img_001167.png", "images/img_001159.png"], "supporting": {"draw_boxes": {"images/img_001159.png": [[131, 118, 103, 264]]}}} +{"q_idx": "detect_000601", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'stylized F22 fighter jet with white and red color'.\n It has 1 pink and maroon striped hot air balloon to its right-front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[220.0, 382.0, 351.0, 482.0]], "image_paths": ["images/img_000894.png", "images/img_000893.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000602", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'three balls snowman'.\n It has 1 police car to its behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[590.0, 530.0, 631.0, 597.0]], "image_paths": ["images/img_000782.png", "images/img_000784.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000603", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'green camouflage helicopter'.\n It has 1 dark grey helicopter to its behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[650.0, 347.0, 746.0, 397.0]], "image_paths": ["images/img_000999.png", "images/img_001007.png", "images/img_001009.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000604", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'man wearing gray windbreaker'.\n It has 1 man wearing gray windbreaker and 1 bus stop with advertisement poster to its left-behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[639.0, 367.0, 698.0, 592.0]], "image_paths": ["images/img_000018.png", "images/img_000026.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000605", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'tombstone without cross'.\n It has 1 beige upright tombstone with cross to its left in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[499.0, 333.0, 532.0, 378.0]], "image_paths": ["images/img_000866.png", "images/img_000857.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000606", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'man holding box'.\n It has 1 man holding box and 1 white tombstone to its left-behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[600.0, 533.0, 636.0, 617.0]], "image_paths": ["images/img_000845.png", "images/img_000844.png", "images/img_000841.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000607", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'three balls snowman'.\n It has 1 man wearing green hood to its left-front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[379.0, 323.0, 438.0, 423.0]], "image_paths": ["images/img_000800.png", "images/img_000794.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000608", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'person wearing blue patterned shirt'.\n It has 1 tree to its right-front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[312.0, 335.0, 367.0, 488.0]], "image_paths": ["images/img_000714.png", "images/img_000716.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000609", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'person with black hat'.\n It has 1 lifeguard tower to its right-behind in view 2. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[614.0, 245.0, 661.0, 385.0]], "image_paths": ["images/img_000682.png", "images/img_000686.png", "images/img_000687.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000610", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'tombstone with pointed top'.\n It has 1 tombstone with shoulder top to its front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[656.0, 333.0, 715.0, 448.0]], "image_paths": ["images/img_001198.png", "images/img_001189.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000611", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'stylized F22 fighter jet with white and red color'.\n It has 1 stylized F22 fighter jet with white and red color to its left-behind, and 1 black and white striped hot air balloon to its left-front in view 1. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[664.0, 485.0, 816.0, 617.0]], "image_paths": ["images/img_000910.png", "images/img_000905.png", "images/img_000909.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000612", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'beige sand helicopter'.\n It has 1 red and white helicopter to its right-behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[552.0, 265.0, 790.0, 403.0]], "image_paths": ["images/img_001004.png", "images/img_000997.png", "images/img_000996.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000613", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'stylized F22 fighter jet with white and red color'.\n It has 1 F-16 fighter jet and 1 stylized F22 fighter jet with white and red color to its behind, and 1 F-16 fighter jet to its left-behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[507.0, 638.0, 631.0, 745.0]], "image_paths": ["images/img_000301.png", "images/img_000303.png", "images/img_000294.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000614", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'dark grey helicopter'.\n It has 1 dark grey helicopter and 1 red and white helicopter to its front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[819.0, 435.0, 938.0, 515.0]], "image_paths": ["images/img_000963.png", "images/img_000972.png", "images/img_000973.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000615", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'man wearing gray hat'.\n It has 1 man wearing gray hat to its behind, and 1 red scarf snowman to its left-behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[589.0, 238.0, 677.0, 533.0]], "image_paths": ["images/img_000735.png", "images/img_000728.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000616", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'man wearing black coat'.\n It has 1 bus stop without enclosure to its right-behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[355.0, 392.0, 417.0, 635.0]], "image_paths": ["images/img_000041.png", "images/img_000039.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000617", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'tombstone without cross'.\n It has 1 black upright tombstone with cross to its front, and 1 black upright tombstone with cross to its right-front in view 3. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[356.0, 625.0, 435.0, 738.0]], "image_paths": ["images/img_001149.png", "images/img_001147.png", "images/img_001148.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000618", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'stylized F22 fighter jet with white and red color'.\n It has 1 mirage fighter jet to its behind, and 1 mirage fighter jet to its right-behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[311.0, 320.0, 435.0, 393.0]], "image_paths": ["images/img_000333.png", "images/img_000324.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000619", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'man wearing brown coat'.\n It has 1 man wearing brown coat to its behind, and 1 bus stop with advertisement poster to its left-front in view 2. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[436.0, 0.0, 634.0, 745.0]], "image_paths": ["images/img_000011.png", "images/img_000009.png", "images/img_000001.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000620", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'white flat tombstone with cross'.\n It has 1 gray colonnade with circular top to its right-behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[524.0, 550.0, 614.0, 662.0]], "image_paths": ["images/img_000829.png", "images/img_000830.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000621", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'harrier jumpjet fighter jet'.\n It has 1 stylized F22 fighter jet with grey and cyan color to its front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[79.0, 447.0, 219.0, 520.0]], "image_paths": ["images/img_000935.png", "images/img_000941.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000622", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'airplane with light blue livery'.\n It has 1 airplane with red livery to its right-front, and 1 airplane with light blue livery to its right in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[22.0, 223.0, 474.0, 397.0]], "image_paths": ["images/img_000310.png", "images/img_000319.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000623", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'airplane with white and orange livery'.\n It has 1 airplane with red livery to its behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[299.0, 505.0, 491.0, 680.0]], "image_paths": ["images/img_000260.png", "images/img_000261.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000624", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'black buttons snowman'.\n It has 1 black buttons snowman and 1 three balls snowman to its right in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[463.0, 223.0, 535.0, 335.0]], "image_paths": ["images/img_000767.png", "images/img_000760.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000625", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'surfboard'.\n It has 1 surfboard and 1 person with tattoos to its right-behind, and 1 person with tattoos to its right-front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[301.0, 393.0, 378.0, 630.0]], "image_paths": ["images/img_000650.png", "images/img_000649.png", "images/img_000643.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000626", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'woman in gray clothing'.\n It has 1 woman in pink clothing to its behind in view 1. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[495.0, 345.0, 536.0, 473.0]], "image_paths": ["images/img_001049.png", "images/img_001052.png", "images/img_001057.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000627", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'man wearing santa costume'.\n It has 1 man wearing gray hat to its front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[399.0, 267.0, 466.0, 458.0]], "image_paths": ["images/img_000033.png", "images/img_000040.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000628", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'person saluting'.\n It has 1 person saluting to its front, and 1 pink flamingo swim ring to its right-front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[261.0, 372.0, 314.0, 502.0]], "image_paths": ["images/img_000702.png", "images/img_000703.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000629", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'lamp'.\n It has 1 stone steps to its left-front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[357.0, 303.0, 386.0, 348.0]], "image_paths": ["images/img_000443.png", "images/img_000435.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000630", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'bench made of white painted wood'.\n It has 3 beach balls and 1 bench made of white painted wood to its front, 1 beach ball to its right-front, 1 beach ball to its right-behind, 1 beach ball to its left-front, and 1 beach ball to its left-behind in view 2. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[361.0, 412.0, 642.0, 652.0]], "image_paths": ["images/img_000599.png", "images/img_000607.png", "images/img_000598.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000631", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'black tombstone'.\n It has 1 white temple to its left-behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[366.0, 393.0, 395.0, 473.0]], "image_paths": ["images/img_000811.png", "images/img_000805.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000632", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'airplane with white and orange livery'.\n It has 1 airplane with white and orange livery to its right-front, 1 airplane with yellow livery to its front, and 1 airplane with yellow livery to its right in view 3. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[21.0, 223.0, 474.0, 397.0]], "image_paths": ["images/img_000292.png", "images/img_000295.png", "images/img_000301.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000633", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'woman wearing underwear'.\n It has 1 woman in pink clothing to its front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[451.0, 433.0, 496.0, 593.0]], "image_paths": ["images/img_001051.png", "images/img_001052.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000634", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'stylized F22 fighter jet with grey and cyan color'.\n It has 1 stylized F22 fighter jet with grey and cyan color and 1 cream and black striped hot air balloon to its left-front in view 1. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[0.0, 260.0, 182.0, 403.0]], "image_paths": ["images/img_000928.png", "images/img_000918.png", "images/img_000920.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000635", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'man holding box'.\n It has 1 man holding box to its right-front, 1 flat tombstone to its right, and 1 flat tombstone to its front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[445.0, 325.0, 471.0, 403.0]], "image_paths": ["images/img_000846.png", "images/img_000847.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000636", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'green camouflage helicopter'.\n It has 1 green camouflage helicopter to its right, 1 dark grey helicopter to its behind, and 1 green camouflage helicopter to its left in view 2. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[529.0, 342.0, 614.0, 398.0]], "image_paths": ["images/img_000991.png", "images/img_000988.png", "images/img_000989.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000637", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'black upright tombstone with cross'.\n It has 1 tombstone without cross to its front, and 1 black upright tombstone with cross to its left in view 2. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[550.0, 210.0, 594.0, 342.0]], "image_paths": ["images/img_001140.png", "images/img_001149.png", "images/img_001148.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000638", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'beige sand helicopter'.\n It has 1 white and green striped hot air balloon to its left-behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[253.0, 577.0, 489.0, 688.0]], "image_paths": ["images/img_001024.png", "images/img_001026.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000639", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'man wearing a dark kurta and white pants'.\n It has 1 man wearing a dark kurta and white pants to its behind, and 1 colonnade with rectangular top to its left-behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[660.0, 427.0, 690.0, 508.0]], "image_paths": ["images/img_000848.png", "images/img_000839.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000640", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'airplane with light blue livery'.\n It has 1 F-16 fighter jet to its left-front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[481.0, 343.0, 595.0, 423.0]], "image_paths": ["images/img_000309.png", "images/img_000319.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000641", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'airplane with white and orange livery'.\n It has 1 F-22 fighter jet with gray color and 1 airplane with white and orange livery to its left-front, and 1 airplane with white and orange livery to its left in view 3. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[154.0, 257.0, 552.0, 397.0]], "image_paths": ["images/img_000277.png", "images/img_000290.png", "images/img_000285.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000642", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'mirage fighter jet'.\n It has 1 F-22 fighter jet with gray color to its right-front in view 3. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[212.0, 332.0, 355.0, 390.0]], "image_paths": ["images/img_000298.png", "images/img_000295.png", "images/img_000304.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000643", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'man wearing gray windbreaker'.\n It has 1 man wearing black coat and 1 man wearing gray windbreaker to its right-behind, 1 man wearing black coat and 1 man wearing gray windbreaker to its left-behind, and 1 man wearing gray windbreaker to its left in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[839.0, 2.0, 1000.0, 915.0]], "image_paths": ["images/img_000026.png", "images/img_000019.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000644", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'white tombstone'.\n It has 1 man in standing posture to its behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[431.0, 338.0, 546.0, 552.0]], "image_paths": ["images/img_000806.png", "images/img_000815.png", "images/img_000812.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000645", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'airplane with yellow livery'.\n It has 1 airplane with yellow livery to its left-behind, 1 airplane with white and orange livery to its front, 1 airplane with yellow livery to its left, and 1 airplane with white and orange livery to its left-front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[181.0, 247.0, 522.0, 395.0]], "image_paths": ["images/img_000219.png", "images/img_000225.png", "images/img_000227.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000646", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'airplane with light blue livery'.\n It has 1 airplane with white and green livery and 1 airplane with light blue livery to its left-front, 1 airplane with white and green livery to its right-front, and 1 airplane with light blue livery to its left in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[200.0, 282.0, 486.0, 397.0]], "image_paths": ["images/img_000203.png", "images/img_000194.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000647", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'stylized F22 fighter jet with camoflague color'.\n It has 1 F-16 fighter jet to its left-behind, and 1 F-16 fighter jet to its behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[223.0, 323.0, 350.0, 393.0]], "image_paths": ["images/img_000316.png", "images/img_000307.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000648", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'trolley'.\n It has 1 trolley to its left-behind, and 1 truck with white and black cab to its left-front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[210.0, 345.0, 244.0, 405.0]], "image_paths": ["images/img_000428.png", "images/img_000426.png", "images/img_000423.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000649", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'white upright tombstone with cross'.\n It has 1 black flat tombstone with cross to its behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[599.0, 322.0, 624.0, 393.0]], "image_paths": ["images/img_000829.png", "images/img_000827.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000650", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'man wearing brown coat'.\n It has 1 man wearing orange coat to its front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[580.0, 285.0, 624.0, 467.0]], "image_paths": ["images/img_000008.png", "images/img_000002.png", "images/img_000003.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000651", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'man wearing brown coat'.\n It has 1 man wearing black coat to its right in view 3. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[691.0, 345.0, 743.0, 535.0]], "image_paths": ["images/img_000072.png", "images/img_000066.png", "images/img_000058.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000652", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'worker'.\n It has 1 worker to its right, and 1 forklift to its left-behind in view 3. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[517.0, 258.0, 531.0, 323.0]], "image_paths": ["images/img_000409.png", "images/img_000410.png", "images/img_000411.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000653", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'airplane with red livery'.\n It has 2 airplane with yellow liveries to its right-front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[580.0, 567.0, 774.0, 657.0]], "image_paths": ["images/img_000214.png", "images/img_000212.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000654", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'black tombstone'.\n It has 1 man in gray suit and tie to its left-front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[652.0, 300.0, 726.0, 417.0]], "image_paths": ["images/img_001172.png", "images/img_001180.png", "images/img_001178.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000655", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'mirage fighter jet'.\n It has 1 harrier jumpjet fighter jet and 1 mirage fighter jet to its front, 2 harrier jumpjet fighter jets to its left, and 1 harrier jumpjet fighter jet to its left-front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[383.0, 308.0, 585.0, 395.0]], "image_paths": ["images/img_000937.png", "images/img_000944.png", "images/img_000930.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000656", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'beige sand helicopter'.\n It has 1 olive green helicopter to its front, and 1 olive green helicopter to its behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[14.0, 277.0, 406.0, 405.0]], "image_paths": ["images/img_001019.png", "images/img_001026.png", "images/img_001018.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000657", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'green camouflage helicopter'.\n It has 1 green camouflage helicopter and 1 black and white striped hot air balloon to its right-behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[884.0, 308.0, 1000.0, 385.0]], "image_paths": ["images/img_000999.png", "images/img_001005.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000658", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'woman wearing black dress'.\n It has 1 woman in pink clothing to its right in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[486.0, 307.0, 529.0, 445.0]], "image_paths": ["images/img_001120.png", "images/img_001113.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000659", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'person saluting'.\n It has 1 person saluting and 1 beach ball to its right-behind, and 1 beach ball to its left-behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[621.0, 460.0, 657.0, 602.0]], "image_paths": ["images/img_000682.png", "images/img_000684.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000660", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'F-16 fighter jet'.\n It has 1 F-16 fighter jet to its behind, 1 F-22 fighter jet with gray color to its front, and 1 F-22 fighter jet with gray color to its right-behind in view 3. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[139.0, 272.0, 480.0, 398.0]], "image_paths": ["images/img_000903.png", "images/img_000912.png", "images/img_000911.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000661", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'woman wearing high boots and purple clothing'.\n It has 2 woman in beige clothings to its left in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[369.0, 490.0, 427.0, 600.0]], "image_paths": ["images/img_001077.png", "images/img_001083.png", "images/img_001087.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000662", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'green wooden bench with armrests'.\n It has 1 child in gray clothing to its left-behind in view 2. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[249.0, 472.0, 430.0, 703.0]], "image_paths": ["images/img_000634.png", "images/img_000639.png", "images/img_000641.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000663", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'bench made of stone'.\n It has 1 green wooden bench with armrests to its right in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[575.0, 333.0, 654.0, 425.0]], "image_paths": ["images/img_000615.png", "images/img_000624.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000664", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'man wearing santa costume'.\n It has 1 man wearing santa costume to its behind, and 2 man wearing green hoods to its right-behind in view 3. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[297.0, 365.0, 369.0, 598.0]], "image_paths": ["images/img_000051.png", "images/img_000052.png", "images/img_000050.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000665", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'surfboard'.\n It has 1 surfboard to its behind, 1 surfboard to its right-front, and 1 surfboard and 1 gray umbrella to its right-behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[311.0, 227.0, 378.0, 455.0]], "image_paths": ["images/img_000671.png", "images/img_000669.png", "images/img_000668.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000666", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'man wearing santa costume'.\n It has 3 snowman with red gloves and black buttonses to its left-front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[489.0, 223.0, 526.0, 310.0]], "image_paths": ["images/img_000748.png", "images/img_000749.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000667", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'worker'.\n It has 1 worker to its right-behind, 1 worker to its front, and 1 truck with blue and white cab to its left-front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[468.0, 268.0, 487.0, 333.0]], "image_paths": ["images/img_000424.png", "images/img_000422.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000668", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'black upright tombstone with cross'.\n It has 1 beige tombstone to its front in view 3. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[604.0, 398.0, 630.0, 470.0]], "image_paths": ["images/img_000840.png", "images/img_000847.png", "images/img_000845.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000669", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'tombstone with pointed top'.\n It has 1 white upright tombstone with cross to its behind, and 1 white upright tombstone with cross to its left-front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[393.0, 235.0, 450.0, 347.0]], "image_paths": ["images/img_001204.png", "images/img_001202.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000670", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'fan'.\n It has 1 stone steps to its left-front in view 2. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[591.0, 767.0, 651.0, 862.0]], "image_paths": ["images/img_000422.png", "images/img_000425.png", "images/img_000427.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000671", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'pink flamingo swim ring'.\n It has 1 gray umbrella to its left in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[494.0, 622.0, 557.0, 692.0]], "image_paths": ["images/img_000716.png", "images/img_000715.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000672", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'surfboard'.\n It has 1 swim ring and 1 surfboard to its right-behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[407.0, 182.0, 450.0, 418.0]], "image_paths": ["images/img_000649.png", "images/img_000651.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000673", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'white upright tombstone with cross'.\n It has 1 white temple to its right-front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[620.0, 390.0, 644.0, 473.0]], "image_paths": ["images/img_000811.png", "images/img_000813.png", "images/img_000806.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000674", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'green camouflage helicopter'.\n It has 1 olive green helicopter to its right-behind, and 1 green camouflage helicopter to its right in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[201.0, 403.0, 314.0, 463.0]], "image_paths": ["images/img_000987.png", "images/img_000980.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000675", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'white flat tombstone with cross'.\n It has 1 beige tombstone to its left-front, and 1 beige tombstone to its left in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[107.0, 363.0, 439.0, 718.0]], "image_paths": ["images/img_001152.png", "images/img_001145.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000676", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'harrier jumpjet fighter jet'.\n It has 1 harrier jumpjet fighter jet and 1 mirage fighter jet to its right in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[656.0, 340.0, 731.0, 407.0]], "image_paths": ["images/img_000936.png", "images/img_000944.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000677", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'airplane with white and green livery'.\n It has 2 airplane with white and green liveries to its right-front, and 1 airplane with red livery to its right in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[824.0, 280.0, 1000.0, 392.0]], "image_paths": ["images/img_000210.png", "images/img_000218.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000678", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'worker'.\n It has 2 workers to its left-front, and 1 red truck to its right-front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[583.0, 810.0, 621.0, 918.0]], "image_paths": ["images/img_000422.png", "images/img_000423.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000679", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'bench made of white painted wood'.\n It has 2 basketballs to its left-behind, and 1 basketball to its right-behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[304.0, 313.0, 407.0, 443.0]], "image_paths": ["images/img_000594.png", "images/img_000601.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000680", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'olive green helicopter'.\n It has 1 olive green helicopter to its left-behind, and 1 red and white helicopter to its right-behind in view 3. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[268.0, 285.0, 370.0, 402.0]], "image_paths": ["images/img_000964.png", "images/img_000965.png", "images/img_000971.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000681", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'woman wearing underwear'.\n It has 1 woman in gray clothing and 1 woman wearing underwear to its behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[385.0, 408.0, 419.0, 563.0]], "image_paths": ["images/img_001115.png", "images/img_001121.png", "images/img_001118.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000682", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'man wearing a green hoodie and orange shoes'.\n It has 1 temple to its left-front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[485.0, 265.0, 516.0, 362.0]], "image_paths": ["images/img_000871.png", "images/img_000879.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000683", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'black upright tombstone with cross'.\n It has 1 shirtless man to its right-front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[641.0, 220.0, 677.0, 345.0]], "image_paths": ["images/img_001204.png", "images/img_001203.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000684", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'concrete mixer'.\n It has 1 concrete mixer to its right-front, 2 concrete mixers to its right-behind, and 1 red truck to its right in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[934.0, 342.0, 1000.0, 488.0]], "image_paths": ["images/img_000437.png", "images/img_000444.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000685", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'tombstone with pointed top'.\n It has 2 white upright tombstone with crosses to its front in view 3. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[305.0, 228.0, 366.0, 332.0]], "image_paths": ["images/img_001135.png", "images/img_001136.png", "images/img_001130.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000686", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'red painted metal bench with armrests'.\n It has 1 child in gray clothing to its right-behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[160.0, 603.0, 411.0, 722.0]], "image_paths": ["images/img_000586.png", "images/img_000585.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000687", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'woman in pink clothing'.\n It has 1 woman wearing black dress to its left in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[388.0, 357.0, 443.0, 472.0]], "image_paths": ["images/img_001096.png", "images/img_001102.png", "images/img_001092.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000688", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'stylized F22 fighter jet with camoflague color'.\n It has 1 airplane with yellow livery to its left-behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[825.0, 545.0, 924.0, 622.0]], "image_paths": ["images/img_000317.png", "images/img_000318.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000689", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'airplane with red livery'.\n It has 2 airplane with white and orange liveries and 2 airplane with red liveries to its right-behind, and 1 airplane with red livery to its right in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[356.0, 265.0, 564.0, 390.0]], "image_paths": ["images/img_000235.png", "images/img_000243.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000690", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'white flat tombstone with cross'.\n It has 1 beige tombstone to its left in view 3. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[366.0, 288.0, 486.0, 432.0]], "image_paths": ["images/img_001164.png", "images/img_001169.png", "images/img_001157.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000691", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'airplane with yellow livery'.\n It has 1 airplane with white and green livery to its right in view 2. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[517.0, 167.0, 865.0, 395.0]], "image_paths": ["images/img_000246.png", "images/img_000239.png", "images/img_000234.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000692", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'man wearing green hood'.\n It has 1 bus stop without advertisement poster to its right-behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[615.0, 365.0, 686.0, 600.0]], "image_paths": ["images/img_000058.png", "images/img_000067.png", "images/img_000066.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000693", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'pink flamingo swim ring'.\n It has 1 lifeguard tower to its right in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[594.0, 393.0, 646.0, 445.0]], "image_paths": ["images/img_000682.png", "images/img_000681.png", "images/img_000685.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000694", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'white flat tombstone with cross'.\n It has 1 beige upright tombstone with cross to its left-behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[0.0, 538.0, 56.0, 720.0]], "image_paths": ["images/img_000864.png", "images/img_000858.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000695", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'green camouflage helicopter'.\n It has 1 green and black striped hot air balloon to its right-behind in view 3. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[461.0, 313.0, 540.0, 367.0]], "image_paths": ["images/img_001031.png", "images/img_001041.png", "images/img_001034.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000696", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'flat tombstone'.\n It has 1 tombstone with shoulder top to its left in view 2. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[699.0, 427.0, 833.0, 578.0]], "image_paths": ["images/img_001177.png", "images/img_001172.png", "images/img_001181.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000697", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'green wooden bench with armrests'.\n It has 1 basketball to its left-behind, 2 basketballs to its right-behind, 3 basketballs to its right-front, 2 basketballs to its left-front, and 1 basketball to its front in view 2. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[383.0, 552.0, 591.0, 705.0]], "image_paths": ["images/img_000560.png", "images/img_000570.png", "images/img_000572.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000698", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'worker'.\n It has 1 worker to its behind, and 1 worker and 1 white stack of sandbags to its left-behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[329.0, 718.0, 360.0, 807.0]], "image_paths": ["images/img_000458.png", "images/img_000456.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000699", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'blue painted metal bench without armrests'.\n It has 2 soccer balls to its right-front, 1 soccer ball to its left-behind, and 1 soccer ball to its left-front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[297.0, 278.0, 472.0, 447.0]], "image_paths": ["images/img_000584.png", "images/img_000583.png", "images/img_000585.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000700", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'airplane with yellow livery'.\n It has 1 mirage fighter jet to its left-behind, and 1 mirage fighter jet to its right-behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[840.0, 182.0, 1000.0, 353.0]], "image_paths": ["images/img_000326.png", "images/img_000334.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000701", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'lamp'.\n It has 2 lamps to its right, 1 worker wearing red hat and white shirt to its right-behind, and 1 worker wearing red hat and white shirt to its left in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[271.0, 692.0, 316.0, 760.0]], "image_paths": ["images/img_000479.png", "images/img_000478.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000702", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'woman in gray clothing'.\n It has 2 woman in gray clothings to its left, and 1 woman wearing black dress to its left-behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[0.0, 57.0, 150.0, 772.0]], "image_paths": ["images/img_001070.png", "images/img_001063.png", "images/img_001068.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000703", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'airplane with white and green livery'.\n It has 1 airplane with light blue livery to its front in view 1. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[331.0, 378.0, 449.0, 423.0]], "image_paths": ["images/img_000193.png", "images/img_000191.png", "images/img_000199.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000704", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'airplane with yellow livery'.\n It has 2 airplane with yellow liveries to its behind, 1 airplane with yellow livery and 1 airplane with white and orange livery to its left, and 1 airplane with yellow livery to its left-behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[892.0, 323.0, 1000.0, 393.0]], "image_paths": ["images/img_000249.png", "images/img_000257.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000705", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'man wearing orange coat'.\n It has 1 bus stop with advertisement poster to its right-behind in view 2. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[328.0, 250.0, 385.0, 442.0]], "image_paths": ["images/img_000013.png", "images/img_000002.png", "images/img_000007.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000706", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'red gloves snowman'.\n It has 1 man wearing orange coat and 1 red gloves snowman to its right-front, and 1 red gloves snowman to its right in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[366.0, 160.0, 596.0, 565.0]], "image_paths": ["images/img_000782.png", "images/img_000774.png", "images/img_000778.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000707", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'concrete mixer'.\n It has 1 concrete mixer and 1 truck with white and black cab to its right-front, and 2 concrete mixers to its right-behind in view 1. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[499.0, 352.0, 583.0, 462.0]], "image_paths": ["images/img_000445.png", "images/img_000440.png", "images/img_000436.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000708", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'bench made of natural wood'.\n It has 1 child in gray clothing to its left-front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[0.0, 353.0, 221.0, 657.0]], "image_paths": ["images/img_000564.png", "images/img_000571.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000709", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'white upright tombstone with cross'.\n It has 2 black upright tombstone with crosses to its behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[720.0, 592.0, 784.0, 715.0]], "image_paths": ["images/img_001169.png", "images/img_001160.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000710", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'olive green helicopter'.\n It has 1 white and green striped hot air balloon to its left-behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[276.0, 338.0, 306.0, 400.0]], "image_paths": ["images/img_001019.png", "images/img_001026.png", "images/img_001012.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000711", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'man wearing orange coat'.\n It has 1 man wearing gray windbreaker to its behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[62.0, 0.0, 395.0, 760.0]], "image_paths": ["images/img_000044.png", "images/img_000051.png", "images/img_000053.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000712", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'man wearing black coat'.\n It has 1 man wearing black coat and 2 man wearing green hoods to its right-behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[461.0, 457.0, 531.0, 712.0]], "image_paths": ["images/img_000057.png", "images/img_000055.png", "images/img_000044.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000713", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'airplane with white and green livery'.\n It has 1 airplane with light blue livery to its front in view 3. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[792.0, 412.0, 936.0, 515.0]], "image_paths": ["images/img_000256.png", "images/img_000261.png", "images/img_000251.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000714", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'man wearing santa costume'.\n It has 1 red buttons snowman to its left-behind, and 1 red buttons snowman to its left-front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[664.0, 417.0, 699.0, 510.0]], "image_paths": ["images/img_000791.png", "images/img_000801.png", "images/img_000792.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000715", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'surfboard'.\n It has 1 surfboard to its behind, 1 person and 1 surfboard to its right-front, and 1 surfboard to its right-behind in view 3. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[509.0, 495.0, 575.0, 755.0]], "image_paths": ["images/img_000666.png", "images/img_000667.png", "images/img_000668.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000716", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'worker wearing red hat and white shirt'.\n It has 1 worker wearing red hat and white shirt and 1 red truck to its left-front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[657.0, 710.0, 688.0, 798.0]], "image_paths": ["images/img_000478.png", "images/img_000479.png", "images/img_000477.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000717", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'airplane with yellow livery'.\n It has 1 airplane with yellow livery to its behind, 1 airplane with white and orange livery to its right-front, 1 airplane with white and orange livery to its front, and 1 airplane with yellow livery to its right in view 2. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[631.0, 377.0, 767.0, 467.0]], "image_paths": ["images/img_000223.png", "images/img_000230.png", "images/img_000229.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000718", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'F-22 fighter jet with gray color'.\n It has 1 F-22 fighter jet with gray color to its left-front, and 1 green and yellow striped hot air balloon to its left-behind in view 2. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[224.0, 383.0, 334.0, 460.0]], "image_paths": ["images/img_000913.png", "images/img_000912.png", "images/img_000911.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000719", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'woman wearing underwear'.\n It has 1 woman wearing underwear to its front, and 1 vendor cart with red and yellow roof to its left-front in view 3. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[443.0, 7.0, 547.0, 673.0]], "image_paths": ["images/img_001123.png", "images/img_001114.png", "images/img_001120.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000720", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'tombstone with pointed top'.\n It has 2 white upright tombstone with crosses to its front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[305.0, 228.0, 366.0, 332.0]], "image_paths": ["images/img_001136.png", "images/img_001130.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000721", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'woman wearing underwear'.\n It has 1 woman wearing black hat to its left-front, 1 woman wearing black hat to its front, and 1 woman wearing black hat and 1 woman wearing underwear to its left-behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[446.0, 322.0, 475.0, 468.0]], "image_paths": ["images/img_001088.png", "images/img_001090.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000722", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'bench made of black metal'.\n It has 1 child in green and brown clothing to its right-front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[172.0, 587.0, 367.0, 718.0]], "image_paths": ["images/img_000605.png", "images/img_000606.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000723", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'tombstone with pointed top'.\n It has 1 tombstone with pointed top to its front, and 1 man in blue suit and tie to its right-front in view 3. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[774.0, 420.0, 844.0, 543.0]], "image_paths": ["images/img_001176.png", "images/img_001187.png", "images/img_001178.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000724", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'F-22 fighter jet with gray color'.\n It has 2 mirage fighter jets to its left-front in view 1. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[660.0, 307.0, 914.0, 403.0]], "image_paths": ["images/img_000927.png", "images/img_000922.png", "images/img_000921.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000725", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'bench made of black metal'.\n It has 2 bench made of black metals to its left, and 1 bench made of black metal and 1 green wooden bench with armrests to its left-front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[140.0, 367.0, 310.0, 560.0]], "image_paths": ["images/img_000637.png", "images/img_000639.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000726", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'person wearing blue patterned shirt'.\n It has 2 toy crabs to its right-front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[386.0, 248.0, 420.0, 402.0]], "image_paths": ["images/img_000713.png", "images/img_000715.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000727", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'airplane with yellow livery'.\n It has 1 airplane with white and green livery to its left, 1 airplane with yellow livery to its right-behind, 1 airplane with white and green livery to its left-behind, and 1 airplane with white and green livery to its right in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[160.0, 455.0, 336.0, 577.0]], "image_paths": ["images/img_000218.png", "images/img_000217.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000728", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'trolley'.\n It has 2 trolleys to its behind, 1 trolley to its right-front, and 1 black stack of sandbags to its right-behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[230.0, 743.0, 311.0, 800.0]], "image_paths": ["images/img_000462.png", "images/img_000461.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000729", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'woman in white clothing'.\n It has 1 woman in white clothing to its behind, and 1 vendor cart with pink and white checkered roof to its left-behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[826.0, 63.0, 1000.0, 762.0]], "image_paths": ["images/img_001070.png", "images/img_001063.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000730", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'airplane with white and orange livery'.\n It has 1 airplane with red livery to its behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[28.0, 182.0, 419.0, 395.0]], "image_paths": ["images/img_000257.png", "images/img_000248.png", "images/img_000256.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000731", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'man wearing orange coat'.\n It has 1 man wearing black coat to its front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[705.0, 577.0, 746.0, 687.0]], "image_paths": ["images/img_000740.png", "images/img_000750.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000732", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'woman in pink clothing'.\n It has 1 woman wearing black dress to its front, and 1 woman in pink clothing to its left in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[432.0, 452.0, 469.0, 593.0]], "image_paths": ["images/img_001070.png", "images/img_001072.png", "images/img_001062.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000733", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'person wearing blue patterned shirt'.\n It has 1 pink flamingo swim ring to its right-front in view 3. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[386.0, 248.0, 420.0, 402.0]], "image_paths": ["images/img_000706.png", "images/img_000715.png", "images/img_000714.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000734", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'man wearing gray hat'.\n It has 1 man wearing red jacket to its behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[482.0, 245.0, 540.0, 425.0]], "image_paths": ["images/img_000032.png", "images/img_000041.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000735", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'airplane with red livery'.\n It has 1 harrier jumpjet fighter jet to its right-behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[256.0, 387.0, 390.0, 477.0]], "image_paths": ["images/img_000328.png", "images/img_000321.png", "images/img_000323.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000736", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'red curved metal bench without armrests'.\n It has 1 blue painted metal bench without armrests to its right-behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[436.0, 655.0, 570.0, 818.0]], "image_paths": ["images/img_000587.png", "images/img_000586.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000737", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'woman wearing high boots and purple clothing'.\n It has 1 woman in pink clothing to its behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[514.0, 382.0, 560.0, 472.0]], "image_paths": ["images/img_001102.png", "images/img_001094.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000738", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'lamp'.\n It has 1 lamp to its left, 1 lamp to its right, and 1 forklift to its right-behind in view 2. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[711.0, 350.0, 741.0, 398.0]], "image_paths": ["images/img_000476.png", "images/img_000465.png", "images/img_000475.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000739", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'black upright tombstone with cross'.\n It has 1 man taking off white shorts to its left-behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[641.0, 220.0, 677.0, 345.0]], "image_paths": ["images/img_001203.png", "images/img_001193.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000740", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'man wearing red jacket'.\n It has 1 man wearing brown coat to its behind, and 1 man wearing red jacket to its left-behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[637.0, 412.0, 711.0, 640.0]], "image_paths": ["images/img_000069.png", "images/img_000068.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000741", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'bench made of black metal'.\n It has 2 bench made of black metals and 1 child in purple clothing to its left-front, and 1 bench made of black metal to its front in view 2. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[398.0, 623.0, 594.0, 865.0]], "image_paths": ["images/img_000627.png", "images/img_000640.png", "images/img_000641.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000742", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'snowman with red gloves and black buttons'.\n It has 1 man wearing black coat to its right-behind, 1 snowman with red gloves and black buttons to its right-front, and 1 snowman with red gloves and black buttons to its right in view 3. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[0.0, 7.0, 357.0, 687.0]], "image_paths": ["images/img_000744.png", "images/img_000739.png", "images/img_000747.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000743", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'green wooden bench with armrests'.\n It has 2 soccer balls to its right-front, 1 soccer ball to its left-behind, 1 soccer ball to its right-behind, and 2 soccer balls to its left-front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[241.0, 382.0, 754.0, 663.0]], "image_paths": ["images/img_000628.png", "images/img_000637.png", "images/img_000633.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000744", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'dark grey helicopter'.\n It has 1 dark grey helicopter to its behind, and 1 cream and black striped hot air balloon to its left-behind in view 3. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[291.0, 588.0, 449.0, 687.0]], "image_paths": ["images/img_000975.png", "images/img_000972.png", "images/img_000974.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000745", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'tombstone with round top'.\n It has 1 man in sitting posture to its front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[365.0, 418.0, 404.0, 462.0]], "image_paths": ["images/img_000829.png", "images/img_000823.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000746", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'green camouflage helicopter'.\n It has 1 desert camouflage helicopter and 1 green camouflage helicopter to its left-behind, and 1 green camouflage helicopter to its left in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[270.0, 572.0, 411.0, 685.0]], "image_paths": ["images/img_001041.png", "images/img_001043.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000747", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'airplane with red livery'.\n It has 1 airplane with white and green livery to its front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[60.0, 565.0, 266.0, 683.0]], "image_paths": ["images/img_000193.png", "images/img_000198.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000748", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'woman in white clothing'.\n It has 2 woman in white clothings to its left-behind, and 1 food stall with red and white striped roof to its left in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[516.0, 342.0, 551.0, 465.0]], "image_paths": ["images/img_001056.png", "images/img_001054.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000749", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'striped scarf snowman'.\n It has 1 coupe to its left-behind in view 3. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[334.0, 307.0, 419.0, 402.0]], "image_paths": ["images/img_000725.png", "images/img_000729.png", "images/img_000722.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000750", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'plaid scarf snowman'.\n It has 1 sedan to its behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[0.0, 150.0, 195.0, 723.0]], "image_paths": ["images/img_000788.png", "images/img_000796.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000751", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'woman in beige clothing'.\n It has 1 woman wearing black hat to its left, and 1 woman wearing black hat and 1 vendor cart with wooden roof to its left-behind in view 3. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[346.0, 390.0, 375.0, 525.0]], "image_paths": ["images/img_001083.png", "images/img_001090.png", "images/img_001089.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000752", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'F-22 fighter jet with gray color'.\n It has 1 stylized F22 fighter jet with white and red color to its left-behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[246.0, 352.0, 380.0, 408.0]], "image_paths": ["images/img_000891.png", "images/img_000893.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000753", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'stylized F22 fighter jet with camoflague color'.\n It has 1 stylized F22 fighter jet with white and red color to its front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[0.0, 350.0, 109.0, 392.0]], "image_paths": ["images/img_000281.png", "images/img_000280.png", "images/img_000278.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000754", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'F-22 fighter jet with gray color'.\n It has 1 stylized F22 fighter jet with white and red color to its left in view 3. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[126.0, 485.0, 314.0, 572.0]], "image_paths": ["images/img_000900.png", "images/img_000910.png", "images/img_000905.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000755", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'striped scarf snowman'.\n It has 1 peach supercar to its left-behind in view 1. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[598.0, 518.0, 684.0, 605.0]], "image_paths": ["images/img_000785.png", "images/img_000779.png", "images/img_000783.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000756", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'man wearing a bandana and a denim vest, sitting'.\n It has 1 temple to its right-behind in view 3. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[611.0, 410.0, 640.0, 475.0]], "image_paths": ["images/img_000816.png", "images/img_000817.png", "images/img_000815.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000757", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'person with black hat'.\n It has 1 lifeguard tower to its left-behind in view 3. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[519.0, 237.0, 555.0, 380.0]], "image_paths": ["images/img_000702.png", "images/img_000704.png", "images/img_000695.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000758", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'white upright tombstone with cross'.\n It has 1 white upright tombstone with cross to its front, and 1 man squatting to its left-behind in view 3. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[544.0, 562.0, 585.0, 703.0]], "image_paths": ["images/img_001132.png", "images/img_001134.png", "images/img_001136.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000759", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'person wearing blue patterned shirt'.\n It has 1 gray umbrella to its front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[598.0, 503.0, 651.0, 662.0]], "image_paths": ["images/img_000652.png", "images/img_000654.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000760", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'man wearing red jacket'.\n It has 1 man wearing red jacket to its behind, and 1 SUV to its left-behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[512.0, 247.0, 586.0, 523.0]], "image_paths": ["images/img_000724.png", "images/img_000731.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000761", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'bench made of black metal'.\n It has 1 baseball to its left-front, and 1 bench made of black metal to its left in view 3. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[676.0, 383.0, 830.0, 540.0]], "image_paths": ["images/img_000573.png", "images/img_000569.png", "images/img_000564.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000762", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'worker wearing face mask and red shirt'.\n It has 1 forklift to its right-front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[411.0, 845.0, 450.0, 935.0]], "image_paths": ["images/img_000415.png", "images/img_000414.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000763", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'F-16 fighter jet'.\n It has 1 pink and maroon striped hot air balloon to its right-front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[635.0, 350.0, 757.0, 405.0]], "image_paths": ["images/img_000958.png", "images/img_000949.png", "images/img_000961.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000764", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'airplane with yellow livery'.\n It has 1 harrier jumpjet fighter jet to its left-behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[354.0, 387.0, 469.0, 432.0]], "image_paths": ["images/img_000274.png", "images/img_000266.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000765", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'F-22 fighter jet with gray color'.\n It has 1 F-22 fighter jet with gray color to its front, and 1 red and black striped hot air balloon to its right-front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[512.0, 352.0, 649.0, 408.0]], "image_paths": ["images/img_000954.png", "images/img_000946.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000766", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'white upright tombstone with cross'.\n It has 2 black upright tombstone with crosses to its behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[0.0, 238.0, 77.0, 615.0]], "image_paths": ["images/img_001163.png", "images/img_001158.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000767", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'man wearing orange coat'.\n It has 1 man wearing santa costume to its right-behind, 1 man wearing santa costume to its right, and 1 man wearing orange coat to its front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[590.0, 198.0, 709.0, 532.0]], "image_paths": ["images/img_000760.png", "images/img_000766.png", "images/img_000762.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000768", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'man dressed in white, sitting on a stool and wearing red shoes'.\n It has 1 colonnade with circular top to its behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[535.0, 355.0, 574.0, 418.0]], "image_paths": ["images/img_000879.png", "images/img_000870.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000769", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'toy crab'.\n It has 1 surfboard to its front, and 1 surfboard to its behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[456.0, 398.0, 495.0, 443.0]], "image_paths": ["images/img_000671.png", "images/img_000663.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000770", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'mirage fighter jet'.\n It has 1 harrier jumpjet fighter jet to its left in view 3. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[430.0, 363.0, 500.0, 393.0]], "image_paths": ["images/img_000271.png", "images/img_000270.png", "images/img_000273.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000771", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'white flat tombstone with cross'.\n It has 1 white temple to its left in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[526.0, 353.0, 938.0, 718.0]], "image_paths": ["images/img_000860.png", "images/img_000852.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000772", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'fan'.\n It has 1 black stack of sandbags to its right-front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[233.0, 363.0, 259.0, 418.0]], "image_paths": ["images/img_000459.png", "images/img_000457.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000773", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'person saluting'.\n It has 1 person saluting to its right, and 1 person wearing blue patterned shirt to its right-front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[525.0, 87.0, 695.0, 685.0]], "image_paths": ["images/img_000701.png", "images/img_000695.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000774", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'lamp'.\n It has 2 lamps and 1 white truck with flatbed to its left-front, and 2 lamps to its front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[839.0, 373.0, 914.0, 462.0]], "image_paths": ["images/img_000466.png", "images/img_000473.png", "images/img_000470.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000775", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'worker wearing red hat and white shirt'.\n It has 1 worker wearing red hat and white shirt to its right, and 1 forklift to its left-behind in view 3. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[517.0, 258.0, 531.0, 323.0]], "image_paths": ["images/img_000409.png", "images/img_000410.png", "images/img_000408.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000776", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'airplane with white and green livery'.\n It has 2 airplane with white and green liveries to its right, 1 airplane with white and green livery and 1 airplane with white and orange livery to its right-front, and 1 airplane with white and orange livery to its front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[390.0, 302.0, 593.0, 392.0]], "image_paths": ["images/img_000222.png", "images/img_000228.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000777", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'blue painted metal bench without armrests'.\n It has 1 stone bench without armrests to its right in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[297.0, 278.0, 472.0, 447.0]], "image_paths": ["images/img_000577.png", "images/img_000584.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000778", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'F-22 fighter jet with gray color'.\n It has 3 stylized F22 fighter jet with grey and cyan colors and 1 F-22 fighter jet with gray color to its right-behind, 1 stylized F22 fighter jet with grey and cyan color to its left-behind, and 1 F-22 fighter jet with gray color to its right in view 3. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[555.0, 267.0, 885.0, 405.0]], "image_paths": ["images/img_000949.png", "images/img_000951.png", "images/img_000958.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000779", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'bench made of blue painted metal'.\n It has 1 bench made of black metal and 1 bench made of blue painted metal to its right in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[147.0, 607.0, 396.0, 715.0], [615.0, 545.0, 848.0, 698.0]], "image_paths": ["images/img_000623.png", "images/img_000622.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000780", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'plaid scarf snowman'.\n It has 1 man wearing gray hat to its right-behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[561.0, 352.0, 598.0, 422.0]], "image_paths": ["images/img_000798.png", "images/img_000790.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000781", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'F-16 fighter jet'.\n It has 1 F-16 fighter jet to its left-behind, and 1 white and green striped hot air balloon to its right-front in view 1. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[1.0, 285.0, 399.0, 400.0]], "image_paths": ["images/img_000897.png", "images/img_000896.png", "images/img_000890.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000782", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'man wearing black coat'.\n It has 1 man wearing black coat to its right-front, and 1 bus stop with advertisement poster to its left-front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[282.0, 340.0, 345.0, 580.0]], "image_paths": ["images/img_000025.png", "images/img_000024.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000783", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'stylized F22 fighter jet with white and red color'.\n It has 1 black and white striped hot air balloon to its left-front in view 1. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[191.0, 378.0, 335.0, 488.0]], "image_paths": ["images/img_000939.png", "images/img_000941.png", "images/img_000938.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000784", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'black tombstone'.\n It has 1 man wearing red t-shirt to its right-front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[579.0, 630.0, 635.0, 765.0]], "image_paths": ["images/img_001136.png", "images/img_001138.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000785", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'stylized F22 fighter jet with grey and cyan color'.\n It has 1 F-22 fighter jet with gray color to its left in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[83.0, 563.0, 276.0, 720.0]], "image_paths": ["images/img_000920.png", "images/img_000926.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000786", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'man with pink pants carrying a phone and a shopping bag'.\n It has 1 colonnade with rectangular top to its left-front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[438.0, 315.0, 464.0, 400.0]], "image_paths": ["images/img_000875.png", "images/img_000883.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000787", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'man wearing green hood'.\n It has 1 man wearing red jacket to its right, and 1 man wearing green hood and 1 man wearing red jacket to its right-behind in view 2. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[301.0, 413.0, 378.0, 665.0]], "image_paths": ["images/img_000059.png", "images/img_000066.png", "images/img_000065.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000788", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'dark grey helicopter'.\n It has 1 cream and black striped hot air balloon to its right-front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[386.0, 338.0, 472.0, 403.0]], "image_paths": ["images/img_001026.png", "images/img_001024.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000789", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'woman wearing underwear'.\n It has 1 woman in gray clothing and 1 woman wearing underwear to its behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[385.0, 408.0, 419.0, 563.0]], "image_paths": ["images/img_001115.png", "images/img_001121.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000790", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'airplane with red livery'.\n It has 2 airplane with light blue liveries to its behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[624.0, 360.0, 756.0, 420.0]], "image_paths": ["images/img_000236.png", "images/img_000243.png", "images/img_000241.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000791", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'person with tattoos'.\n It has 1 person with tattoos to its right, and 1 beige umbrella to its left-front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[605.0, 290.0, 649.0, 420.0]], "image_paths": ["images/img_000655.png", "images/img_000653.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000792", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'airplane with yellow livery'.\n It has 1 airplane with red livery to its right, 1 airplane with red livery to its behind, and 1 airplane with red livery to its right-behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[841.0, 182.0, 1000.0, 353.0]], "image_paths": ["images/img_000268.png", "images/img_000275.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000793", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'woman wearing black dress'.\n It has 1 woman wearing underwear to its left-behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[405.0, 362.0, 440.0, 507.0]], "image_paths": ["images/img_001104.png", "images/img_001102.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000794", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'olive green helicopter'.\n It has 1 white and green striped hot air balloon to its left-front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[688.0, 338.0, 716.0, 397.0]], "image_paths": ["images/img_001009.png", "images/img_000999.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000795", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'green camouflage helicopter'.\n It has 1 beige sand helicopter to its left-behind, and 2 green camouflage helicopters to its left in view 3. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[425.0, 322.0, 578.0, 397.0]], "image_paths": ["images/img_000980.png", "images/img_000984.png", "images/img_000987.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000796", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'woman in gray clothing'.\n It has 1 woman in gray clothing to its behind, 1 woman in gray clothing to its front, and 1 food stall with red and white striped roof to its left-front in view 2. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[394.0, 372.0, 439.0, 508.0]], "image_paths": ["images/img_001061.png", "images/img_001065.png", "images/img_001072.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000797", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'bench made of blue painted metal'.\n It has 1 bench made of blue painted metal to its right, and 1 child in blue and white clothing to its right-front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[708.0, 380.0, 1000.0, 662.0]], "image_paths": ["images/img_000621.png", "images/img_000611.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000798", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'green camouflage helicopter'.\n It has 1 desert camouflage helicopter and 1 green camouflage helicopter to its front, and 1 green camouflage helicopter to its right-front in view 2. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[896.0, 307.0, 1000.0, 385.0]], "image_paths": ["images/img_001036.png", "images/img_001042.png", "images/img_001035.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000799", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'stylized F22 fighter jet with camoflague color'.\n It has 1 F-22 fighter jet with gray color to its right-front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[0.0, 350.0, 109.0, 392.0]], "image_paths": ["images/img_000289.png", "images/img_000280.png", "images/img_000279.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000800", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'black buttons snowman'.\n It has 1 man wearing gray hat to its behind, 1 man wearing gray hat to its right-behind, and 1 black buttons snowman to its right in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[422.0, 547.0, 544.0, 665.0]], "image_paths": ["images/img_000764.png", "images/img_000763.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000801", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'worker'.\n It has 1 worker to its front, and 1 forklift to its right-behind in view 3. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[450.0, 263.0, 470.0, 325.0]], "image_paths": ["images/img_000441.png", "images/img_000444.png", "images/img_000437.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000802", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'airplane with red livery'.\n It has 1 harrier jumpjet fighter jet to its left-front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[509.0, 357.0, 623.0, 437.0]], "image_paths": ["images/img_000276.png", "images/img_000266.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000803", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'man wearing red jacket'.\n It has 1 man wearing black coat to its front in view 1. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[486.0, 243.0, 527.0, 413.0]], "image_paths": ["images/img_000044.png", "images/img_000054.png", "images/img_000051.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000804", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'red and white helicopter'.\n It has 1 olive green helicopter to its left in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[259.0, 348.0, 331.0, 422.0]], "image_paths": ["images/img_001009.png", "images/img_001001.png", "images/img_001010.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000805", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'man wearing green hood'.\n It has 1 bus stop with advertisement poster to its right-behind in view 2. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[315.0, 347.0, 374.0, 585.0]], "image_paths": ["images/img_000050.png", "images/img_000047.png", "images/img_000053.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000806", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'olive green helicopter'.\n It has 1 cream and black striped hot air balloon to its left-front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[740.0, 502.0, 816.0, 577.0]], "image_paths": ["images/img_000990.png", "images/img_000980.png", "images/img_000994.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000807", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'person with black hat'.\n It has 1 person saluting to its left in view 2. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[481.0, 205.0, 515.0, 348.0]], "image_paths": ["images/img_000688.png", "images/img_000679.png", "images/img_000686.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000808", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'bench made of stone'.\n It has 1 basketball to its right-behind, and 1 basketball to its left-behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[346.0, 342.0, 425.0, 435.0]], "image_paths": ["images/img_000622.png", "images/img_000615.png", "images/img_000618.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000809", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'no hat snowman'.\n It has 1 man wearing gray hat to its left-behind, and 1 three balls snowman to its left in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[545.0, 313.0, 604.0, 418.0]], "image_paths": ["images/img_000765.png", "images/img_000755.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000810", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'beige upright tombstone with cross'.\n It has 1 man wearing a green hoodie and orange shoes to its right in view 1. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[604.0, 375.0, 634.0, 460.0]], "image_paths": ["images/img_000870.png", "images/img_000883.png", "images/img_000879.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000811", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'woman in beige clothing'.\n It has 1 woman in white clothing to its left in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[389.0, 375.0, 419.0, 508.0]], "image_paths": ["images/img_001088.png", "images/img_001081.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000812", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'woman in pink clothing'.\n It has 1 woman wearing black hat to its left in view 1. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[490.0, 330.0, 521.0, 448.0]], "image_paths": ["images/img_001096.png", "images/img_001100.png", "images/img_001103.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000813", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'snowman with red gloves and black buttons'.\n It has 1 jacket snowman to its behind in view 3. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[557.0, 555.0, 670.0, 665.0]], "image_paths": ["images/img_000797.png", "images/img_000801.png", "images/img_000787.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000814", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'concrete mixer'.\n It has 1 concrete mixer and 1 stone steps to its right-behind in view 3. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[324.0, 335.0, 362.0, 398.0]], "image_paths": ["images/img_000445.png", "images/img_000441.png", "images/img_000435.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000815", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'natural wooden bench with armrests'.\n It has 1 child in purple clothing to its left-behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[145.0, 385.0, 284.0, 538.0]], "image_paths": ["images/img_000564.png", "images/img_000571.png", "images/img_000569.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000816", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'worker'.\n It has 1 worker to its front, and 1 forklift to its right-behind in view 3. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[625.0, 287.0, 645.0, 353.0]], "image_paths": ["images/img_000442.png", "images/img_000445.png", "images/img_000437.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000817", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'airplane with yellow livery'.\n It has 1 airplane with light blue livery to its front, and 1 airplane with light blue livery to its behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[693.0, 482.0, 861.0, 562.0]], "image_paths": ["images/img_000205.png", "images/img_000213.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000818", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'woman in pink clothing'.\n It has 1 woman in beige clothing to its behind in view 1. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[399.0, 413.0, 431.0, 552.0]], "image_paths": ["images/img_001097.png", "images/img_001105.png", "images/img_001103.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000819", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'woman wearing high boots'.\n It has 1 woman in pink clothing to its front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[378.0, 385.0, 425.0, 477.0]], "image_paths": ["images/img_001050.png", "images/img_001057.png", "images/img_001054.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000820", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'worker'.\n It has 1 worker to its behind, and 1 truck with white and black cab to its left-front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[188.0, 472.0, 224.0, 553.0]], "image_paths": ["images/img_000459.png", "images/img_000453.png", "images/img_000456.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000821", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'man wearing red jacket'.\n It has 1 man wearing santa costume to its front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[570.0, 280.0, 619.0, 480.0]], "image_paths": ["images/img_000064.png", "images/img_000072.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000822", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'black tombstone'.\n It has 1 man squatting to its left-behind in view 2. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[245.0, 442.0, 295.0, 578.0]], "image_paths": ["images/img_001136.png", "images/img_001127.png", "images/img_001132.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000823", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'swim ring'.\n It has 1 person to its right-behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[357.0, 420.0, 407.0, 490.0]], "image_paths": ["images/img_000666.png", "images/img_000660.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000824", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'tombstone with pointed top'.\n It has 1 tombstone with pointed top to its front, and 1 man squatting to its right-front in view 2. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[682.0, 588.0, 765.0, 703.0]], "image_paths": ["images/img_001182.png", "images/img_001174.png", "images/img_001184.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000825", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'man wearing brown coat'.\n It has 1 SUV to its left-behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[642.0, 315.0, 672.0, 397.0]], "image_paths": ["images/img_000753.png", "images/img_000743.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000826", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'white upright tombstone with cross'.\n It has 1 black upright tombstone with cross to its behind in view 1. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[461.0, 183.0, 500.0, 305.0]], "image_paths": ["images/img_001159.png", "images/img_001164.png", "images/img_001166.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000827", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'black flat tombstone with cross'.\n It has 1 black flat tombstone with cross and 1 man wearing white thobe to its left-behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[276.0, 490.0, 362.0, 652.0]], "image_paths": ["images/img_001163.png", "images/img_001157.png", "images/img_001166.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000828", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'airplane with white and orange livery'.\n It has 1 airplane with red livery to its front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[693.0, 480.0, 861.0, 560.0]], "image_paths": ["images/img_000232.png", "images/img_000241.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000829", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'swim ring'.\n It has 1 lifeguard tower to its left in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[435.0, 372.0, 485.0, 428.0]], "image_paths": ["images/img_000667.png", "images/img_000660.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000830", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'woman wearing high boots and purple clothing'.\n It has 2 woman wearing black hats to its left in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[369.0, 490.0, 427.0, 600.0]], "image_paths": ["images/img_001077.png", "images/img_001083.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000831", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'airplane with yellow livery'.\n It has 1 airplane with red livery and 1 airplane with yellow livery to its left in view 3. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[62.0, 412.0, 206.0, 515.0]], "image_paths": ["images/img_000259.png", "images/img_000257.png", "images/img_000252.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000832", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'dark grey helicopter'.\n It has 1 green and black striped hot air balloon to its right-behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[199.0, 403.0, 310.0, 463.0]], "image_paths": ["images/img_001040.png", "images/img_001034.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000833", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'beige upright tombstone with cross'.\n It has 1 beige upright tombstone with cross and 1 flat tombstone to its right in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[655.0, 317.0, 695.0, 447.0]], "image_paths": ["images/img_001203.png", "images/img_001193.png", "images/img_001201.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000834", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'bench made of natural wood'.\n It has 3 basketballs to its right-front, 1 basketball to its front, and 1 bench made of natural wood to its right in view 3. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[205.0, 548.0, 301.0, 745.0]], "image_paths": ["images/img_000602.png", "images/img_000604.png", "images/img_000598.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000835", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'F-16 fighter jet'.\n It has 1 green and black striped hot air balloon to its right-behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[674.0, 373.0, 771.0, 453.0]], "image_paths": ["images/img_000952.png", "images/img_000961.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000836", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'person with tattoos'.\n It has 1 beach ball to its right-front in view 1. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[510.0, 225.0, 546.0, 367.0]], "image_paths": ["images/img_000712.png", "images/img_000715.png", "images/img_000720.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000837", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'man wearing green hood'.\n It has 1 taxi to its right-front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[608.0, 273.0, 636.0, 370.0]], "image_paths": ["images/img_000794.png", "images/img_000802.png", "images/img_000799.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000838", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'stylized F22 fighter jet with camoflague color'.\n It has 1 harrier jumpjet fighter jet to its right in view 2. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[98.0, 432.0, 231.0, 538.0]], "image_paths": ["images/img_000906.png", "images/img_000902.png", "images/img_000907.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000839", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'man wearing brown coat'.\n It has 1 man wearing black coat to its front, and 1 man wearing brown coat to its behind in view 3. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[472.0, 267.0, 520.0, 438.0]], "image_paths": ["images/img_000007.png", "images/img_000010.png", "images/img_000001.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000840", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'airplane with red livery'.\n It has 1 stylized F22 fighter jet with camoflague color to its right-front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[564.0, 393.0, 691.0, 440.0]], "image_paths": ["images/img_000308.png", "images/img_000315.png", "images/img_000317.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000841", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'airplane with yellow livery'.\n It has 1 airplane with white and orange livery and 1 airplane with yellow livery to its right in view 1. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[370.0, 375.0, 494.0, 420.0]], "image_paths": ["images/img_000249.png", "images/img_000257.png", "images/img_000256.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000842", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'man wearing a green hoodie and orange shoes'.\n It has 1 beige upright tombstone with cross to its behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[304.0, 407.0, 334.0, 508.0]], "image_paths": ["images/img_000877.png", "images/img_000870.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000843", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'pink flamingo swim ring'.\n It has 1 tree to its left-front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[352.0, 397.0, 404.0, 450.0]], "image_paths": ["images/img_000712.png", "images/img_000718.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000844", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'man wearing brown coat'.\n It has 1 man wearing brown coat to its behind, and 1 man wearing red jacket to its left-behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[406.0, 292.0, 450.0, 470.0]], "image_paths": ["images/img_000006.png", "images/img_000014.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000845", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'man dressed in white, sitting on a stool and wearing red shoes'.\n It has 1 colonnade with open top to its behind in view 1. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[371.0, 405.0, 410.0, 468.0]], "image_paths": ["images/img_000870.png", "images/img_000879.png", "images/img_000877.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000846", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'man wearing brown coat'.\n It has 1 bus stop with advertisement poster to its left-front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[580.0, 285.0, 624.0, 467.0]], "image_paths": ["images/img_000008.png", "images/img_000002.png", "images/img_000012.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000847", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'airplane with yellow livery'.\n It has 2 airplane with yellow liveries to its behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[244.0, 360.0, 376.0, 420.0]], "image_paths": ["images/img_000254.png", "images/img_000247.png", "images/img_000255.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000848", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'tombstone without cross'.\n It has 1 beige tombstone to its left, and 1 beige tombstone to its right in view 2. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[411.0, 247.0, 472.0, 360.0]], "image_paths": ["images/img_001147.png", "images/img_001144.png", "images/img_001150.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000849", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'beige upright tombstone with cross'.\n It has 1 black upright round tombstone to its behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[203.0, 395.0, 260.0, 532.0]], "image_paths": ["images/img_001198.png", "images/img_001192.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000850", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'person with black hat'.\n It has 1 person saluting to its right, and 1 person saluting to its left in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[519.0, 237.0, 555.0, 380.0]], "image_paths": ["images/img_000695.png", "images/img_000704.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000851", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'woman wearing black dress'.\n It has 1 woman in pink clothing to its front in view 1. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[361.0, 390.0, 385.0, 533.0]], "image_paths": ["images/img_001112.png", "images/img_001117.png", "images/img_001118.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000852", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'bench made of red painted metal'.\n It has 1 bench made of stone to its left in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[606.0, 333.0, 708.0, 382.0]], "image_paths": ["images/img_000623.png", "images/img_000615.png", "images/img_000622.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000853", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'olive green helicopter'.\n It has 1 dark grey helicopter to its front in view 1. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[465.0, 315.0, 526.0, 365.0]], "image_paths": ["images/img_000999.png", "images/img_001009.png", "images/img_001008.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000854", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'jacket snowman'.\n It has 1 man wearing red jacket to its right-behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[619.0, 417.0, 665.0, 503.0]], "image_paths": ["images/img_000733.png", "images/img_000723.png", "images/img_000732.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000855", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'man wearing gray windbreaker'.\n It has 1 man wearing black coat and 1 man wearing gray windbreaker to its right-behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[360.0, 280.0, 417.0, 475.0]], "image_paths": ["images/img_000018.png", "images/img_000024.png", "images/img_000021.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000856", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'green camouflage helicopter'.\n It has 2 green camouflage helicopters to its left, and 1 cream and black striped hot air balloon to its left-front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[458.0, 312.0, 537.0, 362.0]], "image_paths": ["images/img_000981.png", "images/img_000988.png", "images/img_000991.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000857", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'mirage fighter jet'.\n It has 1 airplane with white and orange livery to its right-front in view 3. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[164.0, 427.0, 240.0, 455.0]], "image_paths": ["images/img_000302.png", "images/img_000303.png", "images/img_000295.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000858", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'man wearing green hood'.\n It has 1 man wearing gray hat to its front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[244.0, 385.0, 280.0, 480.0]], "image_paths": ["images/img_000796.png", "images/img_000789.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000859", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'man in standing posture'.\n It has 1 white tombstone to its right in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[609.0, 375.0, 636.0, 475.0]], "image_paths": ["images/img_000815.png", "images/img_000805.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000860", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'woman wearing high boots and purple clothing'.\n It has 1 food stall to its left-behind in view 2. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[573.0, 467.0, 608.0, 562.0]], "image_paths": ["images/img_001107.png", "images/img_001094.png", "images/img_001104.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000861", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'stylized F22 fighter jet with white and red color'.\n It has 1 stylized F22 fighter jet with white and red color to its behind, and 1 pink and maroon striped hot air balloon to its left-behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[679.0, 448.0, 826.0, 590.0]], "image_paths": ["images/img_000918.png", "images/img_000928.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000862", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'pink flamingo swim ring'.\n It has 1 beige umbrella to its right in view 1. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[352.0, 397.0, 404.0, 450.0]], "image_paths": ["images/img_000712.png", "images/img_000716.png", "images/img_000718.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000863", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'dark grey helicopter'.\n It has 1 white and green striped hot air balloon to its left-behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[386.0, 338.0, 472.0, 403.0]], "image_paths": ["images/img_001019.png", "images/img_001026.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000864", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'lamp'.\n It has 2 lamps to its front, and 1 forklift to its right-front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[807.0, 403.0, 841.0, 457.0]], "image_paths": ["images/img_000465.png", "images/img_000475.png", "images/img_000474.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000865", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'black tombstone'.\n It has 1 white tombstone to its right in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[378.0, 362.0, 422.0, 438.0]], "image_paths": ["images/img_000851.png", "images/img_000859.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000866", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'airplane with white and green livery'.\n It has 1 airplane with white and green livery to its right, and 1 airplane with white and green livery to its left in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[665.0, 377.0, 782.0, 467.0]], "image_paths": ["images/img_000231.png", "images/img_000223.png", "images/img_000230.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000867", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'woman in white clothing'.\n It has 1 woman in white clothing and 1 woman wearing black dress to its behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[396.0, 342.0, 424.0, 463.0]], "image_paths": ["images/img_001069.png", "images/img_001061.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000868", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'man wearing gray windbreaker'.\n It has 1 man wearing green hood to its front in view 3. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[422.0, 270.0, 463.0, 462.0]], "image_paths": ["images/img_000027.png", "images/img_000028.png", "images/img_000020.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000869", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'man wearing santa costume'.\n It has 1 plaid scarf snowman to its left-behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[586.0, 247.0, 620.0, 347.0]], "image_paths": ["images/img_000778.png", "images/img_000786.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000870", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'mirage fighter jet'.\n It has 1 white and green striped hot air balloon to its right-front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[635.0, 343.0, 769.0, 400.0]], "image_paths": ["images/img_000885.png", "images/img_000891.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000871", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'man wearing brown coat'.\n It has 1 man wearing black coat to its front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[494.0, 250.0, 534.0, 417.0]], "image_paths": ["images/img_000005.png", "images/img_000014.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000872", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'black tombstone'.\n It has 1 pavilion to its left-behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[561.0, 357.0, 639.0, 412.0]], "image_paths": ["images/img_000853.png", "images/img_000861.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000873", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'striped scarf snowman'.\n It has 3 man wearing gray windbreakers to its right-front in view 3. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[559.0, 295.0, 599.0, 387.0]], "image_paths": ["images/img_000779.png", "images/img_000781.png", "images/img_000773.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000874", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'black hat snowman'.\n It has 1 man wearing santa costume to its right-front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[343.0, 298.0, 403.0, 383.0]], "image_paths": ["images/img_000726.png", "images/img_000733.png", "images/img_000734.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000875", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'bench made of blue painted metal'.\n It has 1 green wooden bench with armrests to its left in view 3. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[266.0, 278.0, 434.0, 387.0]], "image_paths": ["images/img_000569.png", "images/img_000568.png", "images/img_000560.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000876", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'airplane with yellow livery'.\n It has 1 F-16 fighter jet to its right-front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[647.0, 482.0, 811.0, 558.0]], "image_paths": ["images/img_000292.png", "images/img_000301.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000877", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'stone bench without armrests'.\n It has 1 red painted metal bench with armrests to its left in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[325.0, 332.0, 412.0, 435.0]], "image_paths": ["images/img_000588.png", "images/img_000581.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000878", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'worker'.\n It has 1 worker to its left, and 1 forklift to its left-behind in view 1. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[347.0, 267.0, 370.0, 325.0]], "image_paths": ["images/img_000436.png", "images/img_000438.png", "images/img_000444.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000879", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'bench made of natural wood'.\n It has 3 tennis balls to its left-behind in view 2. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[286.0, 280.0, 400.0, 392.0]], "image_paths": ["images/img_000566.png", "images/img_000564.png", "images/img_000572.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000880", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'man wearing black coat'.\n It has 1 snowman with black hat and black buttons to its right-behind, and 1 snowman with black hat and black buttons to its right-front in view 1. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[456.0, 215.0, 481.0, 308.0]], "image_paths": ["images/img_000741.png", "images/img_000749.png", "images/img_000748.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000881", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'airplane with red livery'.\n It has 2 airplane with light blue liveries to its behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[757.0, 412.0, 921.0, 515.0]], "image_paths": ["images/img_000236.png", "images/img_000244.png", "images/img_000241.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000882", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'F-16 fighter jet'.\n It has 2 F-22 fighter jet with gray colors to its right in view 3. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[89.0, 595.0, 287.0, 695.0]], "image_paths": ["images/img_000954.png", "images/img_000961.png", "images/img_000948.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000883", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'bench made of black metal'.\n It has 1 basketball to its left-behind, 1 basketball to its right-behind, 1 bench made of black metal to its right, and 1 bench made of black metal to its left in view 3. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[300.0, 288.0, 441.0, 455.0]], "image_paths": ["images/img_000639.png", "images/img_000638.png", "images/img_000632.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000884", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'black upright tombstone with cross'.\n It has 1 black upright tombstone with cross to its left, and 1 shirtless man to its left-behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[399.0, 227.0, 438.0, 348.0]], "image_paths": ["images/img_001147.png", "images/img_001140.png", "images/img_001152.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000885", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'person with black hat'.\n It has 1 person saluting to its left in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[614.0, 245.0, 661.0, 385.0]], "image_paths": ["images/img_000679.png", "images/img_000687.png", "images/img_000686.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000886", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'worker wearing red hat and white shirt'.\n It has 1 forklift to its left-behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[219.0, 357.0, 240.0, 422.0]], "image_paths": ["images/img_000473.png", "images/img_000466.png", "images/img_000476.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000887", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'airplane with yellow livery'.\n It has 1 airplane with yellow livery and 1 airplane with white and orange livery to its left in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[244.0, 360.0, 376.0, 420.0]], "image_paths": ["images/img_000254.png", "images/img_000249.png", "images/img_000255.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000888", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'airplane with light blue livery'.\n It has 1 airplane with light blue livery to its right, and 1 airplane with red livery to its left in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[180.0, 405.0, 338.0, 473.0]], "image_paths": ["images/img_000194.png", "images/img_000200.png", "images/img_000197.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000889", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'jacket snowman'.\n It has 1 man wearing gray windbreaker to its right-behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[315.0, 407.0, 364.0, 487.0]], "image_paths": ["images/img_000751.png", "images/img_000745.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000890", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'olive green helicopter'.\n It has 1 black and white striped hot air balloon to its right-front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[260.0, 397.0, 321.0, 452.0]], "image_paths": ["images/img_000965.png", "images/img_000972.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000891", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'man in sitting posture'.\n It has 1 man holding box to its right in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[364.0, 418.0, 401.0, 480.0]], "image_paths": ["images/img_000860.png", "images/img_000854.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000892", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'man wearing green hood'.\n It has 1 black buttons snowman to its left-behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[280.0, 325.0, 314.0, 428.0]], "image_paths": ["images/img_000761.png", "images/img_000767.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000893", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'harrier jumpjet fighter jet'.\n It has 1 mirage fighter jet to its front, and 1 mirage fighter jet to its behind in view 3. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[733.0, 473.0, 838.0, 572.0]], "image_paths": ["images/img_000940.png", "images/img_000943.png", "images/img_000930.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000894", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'woman wearing underwear'.\n It has 1 woman in white clothing to its front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[564.0, 402.0, 600.0, 557.0]], "image_paths": ["images/img_001057.png", "images/img_001048.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000895", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'man wearing a green hoodie and orange shoes'.\n It has 1 colonnade with circular top to its right-behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[385.0, 383.0, 410.0, 480.0]], "image_paths": ["images/img_000859.png", "images/img_000852.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000896", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'green camouflage helicopter'.\n It has 1 beige sand helicopter to its right in view 1. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[85.0, 435.0, 191.0, 515.0]], "image_paths": ["images/img_000997.png", "images/img_001005.png", "images/img_001004.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000897", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'woman in white clothing'.\n It has 2 woman wearing high bootses and 1 woman in white clothing to its left in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[396.0, 352.0, 443.0, 477.0]], "image_paths": ["images/img_001045.png", "images/img_001051.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000898", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'white upright tombstone with cross'.\n It has 1 beige upright tombstone with cross to its front in view 2. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[263.0, 332.0, 314.0, 455.0]], "image_paths": ["images/img_001181.png", "images/img_001176.png", "images/img_001182.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000899", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'stylized F22 fighter jet with camoflague color'.\n It has 1 mirage fighter jet to its front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[816.0, 487.0, 899.0, 535.0]], "image_paths": ["images/img_000264.png", "images/img_000273.png", "images/img_000271.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000900", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'person with tattoos'.\n It has 1 pink flamingo swim ring to its left in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[400.0, 252.0, 441.0, 387.0]], "image_paths": ["images/img_000712.png", "images/img_000719.png", "images/img_000717.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000901", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'black flat tombstone with cross'.\n It has 1 white flat tombstone with cross and 1 black flat tombstone with cross to its left-behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[429.0, 298.0, 490.0, 412.0]], "image_paths": ["images/img_001157.png", "images/img_001163.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000902", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'F-16 fighter jet'.\n It has 1 F-16 fighter jet to its behind, and 1 airplane with red livery to its left-front in view 3. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[436.0, 363.0, 500.0, 392.0]], "image_paths": ["images/img_000285.png", "images/img_000290.png", "images/img_000278.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000903", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'lamp'.\n It has 1 lamp to its left, and 1 worker wearing red hat and white shirt and 1 lamp to its right in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[350.0, 307.0, 379.0, 352.0]], "image_paths": ["images/img_000466.png", "images/img_000473.png", "images/img_000476.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000904", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'red and white helicopter'.\n It has 1 green camouflage helicopter to its front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[534.0, 365.0, 640.0, 420.0]], "image_paths": ["images/img_000979.png", "images/img_000987.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000905", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'airplane with white and orange livery'.\n It has 1 airplane with yellow livery to its front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[356.0, 372.0, 490.0, 432.0]], "image_paths": ["images/img_000296.png", "images/img_000303.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000906", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'beige sand helicopter'.\n It has 1 red and white helicopter to its behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[233.0, 340.0, 381.0, 397.0]], "image_paths": ["images/img_000971.png", "images/img_000963.png", "images/img_000977.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000907", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'tombstone with pointed top'.\n It has 1 beige upright tombstone with cross to its front in view 2. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[675.0, 475.0, 749.0, 583.0]], "image_paths": ["images/img_001198.png", "images/img_001189.png", "images/img_001199.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000908", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'toy crab'.\n It has 1 blue swim ring to its behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[595.0, 422.0, 642.0, 465.0]], "image_paths": ["images/img_000679.png", "images/img_000688.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000909", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'white upright tombstone with cross'.\n It has 1 shirtless man to its left-front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[705.0, 427.0, 754.0, 563.0]], "image_paths": ["images/img_001140.png", "images/img_001149.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000910", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'airplane with light blue livery'.\n It has 1 airplane with light blue livery to its right, and 1 airplane with red livery to its left in view 3. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[343.0, 348.0, 445.0, 428.0]], "image_paths": ["images/img_000201.png", "images/img_000197.png", "images/img_000194.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000911", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'surfboard'.\n It has 1 lifeguard tower to its left-front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[328.0, 235.0, 381.0, 468.0]], "image_paths": ["images/img_000660.png", "images/img_000667.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000912", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'airplane with red livery'.\n It has 1 F-16 fighter jet to its right-behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[284.0, 372.0, 419.0, 463.0]], "image_paths": ["images/img_000308.png", "images/img_000313.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000913", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'worker'.\n It has 1 worker to its behind, and 1 white stack of sandbags to its right-front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[438.0, 273.0, 458.0, 337.0]], "image_paths": ["images/img_000461.png", "images/img_000453.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000914", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'bench made of natural wood'.\n It has 2 bench made of black metals to its left in view 3. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[449.0, 242.0, 549.0, 362.0]], "image_paths": ["images/img_000572.png", "images/img_000573.png", "images/img_000564.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000915", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'harrier jumpjet fighter jet'.\n It has 1 F-22 fighter jet with gray color to its left-behind in view 1. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[549.0, 358.0, 635.0, 413.0]], "image_paths": ["images/img_000902.png", "images/img_000911.png", "images/img_000909.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000916", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'bench made of black metal'.\n It has 1 beach ball to its left-behind, 1 beach ball to its behind, and 2 bench made of black metals to its left in view 1. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[216.0, 508.0, 338.0, 750.0]], "image_paths": ["images/img_000632.png", "images/img_000635.png", "images/img_000638.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000917", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'white upright tombstone with cross'.\n It has 1 black tombstone to its behind in view 2. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[365.0, 337.0, 386.0, 412.0]], "image_paths": ["images/img_000833.png", "images/img_000820.png", "images/img_000827.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000918", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'man wearing orange coat'.\n It has 1 plaid scarf snowman to its left-behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[345.0, 277.0, 381.0, 383.0]], "image_paths": ["images/img_000784.png", "images/img_000778.png", "images/img_000785.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000919", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'mirage fighter jet'.\n It has 1 white and green striped hot air balloon to its right-front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[785.0, 585.0, 916.0, 720.0]], "image_paths": ["images/img_000893.png", "images/img_000885.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000920", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'airplane with red livery'.\n It has 1 harrier jumpjet fighter jet to its right-behind in view 3. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[534.0, 387.0, 650.0, 432.0]], "image_paths": ["images/img_000331.png", "images/img_000330.png", "images/img_000322.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000921", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'F-22 fighter jet with gray color'.\n It has 1 F-16 fighter jet to its front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[385.0, 363.0, 459.0, 392.0]], "image_paths": ["images/img_000309.png", "images/img_000318.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000922", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'beige tombstone'.\n It has 1 flat tombstone to its front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[588.0, 365.0, 624.0, 400.0]], "image_paths": ["images/img_000835.png", "images/img_000845.png", "images/img_000849.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000923", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'stylized F22 fighter jet with white and red color'.\n It has 1 black and white striped hot air balloon to its right-behind in view 3. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[435.0, 298.0, 544.0, 380.0]], "image_paths": ["images/img_000927.png", "images/img_000925.png", "images/img_000920.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000924", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'F-16 fighter jet'.\n It has 1 F-22 fighter jet with gray color to its front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[659.0, 408.0, 784.0, 463.0]], "image_paths": ["images/img_000892.png", "images/img_000885.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000925", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'mirage fighter jet'.\n It has 1 harrier jumpjet fighter jet to its right-behind in view 2. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[535.0, 335.0, 645.0, 412.0]], "image_paths": ["images/img_000928.png", "images/img_000919.png", "images/img_000927.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000926", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'worker wearing red hat and white shirt'.\n It has 3 lamps to its left in view 3. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[764.0, 350.0, 782.0, 415.0]], "image_paths": ["images/img_000476.png", "images/img_000477.png", "images/img_000466.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000927", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'woman wearing black hat'.\n It has 1 food stall with red and white striped roof to its left-front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[390.0, 382.0, 416.0, 513.0]], "image_paths": ["images/img_001066.png", "images/img_001073.png", "images/img_001069.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000928", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'airplane with white and green livery'.\n It has 2 airplane with yellow liveries to its front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[48.0, 453.0, 215.0, 508.0]], "image_paths": ["images/img_000227.png", "images/img_000221.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000929", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'stone bench without armrests'.\n It has 1 child in blue and white clothing to its behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[575.0, 333.0, 654.0, 425.0]], "image_paths": ["images/img_000615.png", "images/img_000624.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000930", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'airplane with white and green livery'.\n It has 2 airplane with yellow liveries to its front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[636.0, 340.0, 756.0, 428.0]], "image_paths": ["images/img_000230.png", "images/img_000221.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000931", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'man wearing santa costume'.\n It has 2 man wearing green hoods to its behind in view 3. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[343.0, 258.0, 399.0, 457.0]], "image_paths": ["images/img_000054.png", "images/img_000050.png", "images/img_000048.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000932", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'man wearing gray hat'.\n It has 1 bus stop without enclosure to its left-front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[345.0, 265.0, 388.0, 458.0]], "image_paths": ["images/img_000040.png", "images/img_000034.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000933", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'mirage fighter jet'.\n It has 1 airplane with red livery to its right-front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[163.0, 427.0, 240.0, 455.0]], "image_paths": ["images/img_000326.png", "images/img_000332.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000934", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'woman wearing high boots and purple clothing'.\n It has 2 woman in gray clothings to its front in view 2. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[400.0, 392.0, 430.0, 485.0]], "image_paths": ["images/img_001086.png", "images/img_001079.png", "images/img_001085.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000935", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'person with black hat'.\n It has 1 toy crab to its left in view 1. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[694.0, 378.0, 743.0, 530.0]], "image_paths": ["images/img_000679.png", "images/img_000683.png", "images/img_000688.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000936", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'person saluting'.\n It has 1 person saluting and 1 person with black hat to its right in view 1. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[318.0, 367.0, 351.0, 502.0]], "image_paths": ["images/img_000695.png", "images/img_000700.png", "images/img_000701.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000937", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'mirage fighter jet'.\n It has 1 F-16 fighter jet to its front in view 2. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[785.0, 585.0, 916.0, 720.0]], "image_paths": ["images/img_000894.png", "images/img_000885.png", "images/img_000893.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000938", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'black tombstone'.\n It has 1 white temple to its left-behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[366.0, 393.0, 395.0, 473.0]], "image_paths": ["images/img_000811.png", "images/img_000805.png", "images/img_000813.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000939", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'bench made of black metal'.\n It has 1 natural wooden bench with armrests to its right in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[669.0, 505.0, 799.0, 750.0]], "image_paths": ["images/img_000604.png", "images/img_000594.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000940", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'airplane with yellow livery'.\n It has 1 stylized F22 fighter jet with grey and cyan color to its left-front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[354.0, 387.0, 469.0, 432.0]], "image_paths": ["images/img_000274.png", "images/img_000266.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000941", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'white tombstone'.\n It has 1 colonnade with open top to its left-front in view 1. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[588.0, 343.0, 626.0, 407.0]], "image_paths": ["images/img_000806.png", "images/img_000818.png", "images/img_000815.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000942", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'worker'.\n It has 1 worker to its behind, and 1 white stack of sandbags to its left-behind in view 1. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[224.0, 368.0, 251.0, 430.0]], "image_paths": ["images/img_000453.png", "images/img_000455.png", "images/img_000460.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000943", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'lamp'.\n It has 2 lamps to its right, and 1 forklift to its right-front in view 3. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[264.0, 352.0, 294.0, 400.0]], "image_paths": ["images/img_000473.png", "images/img_000478.png", "images/img_000465.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000944", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'green camouflage helicopter'.\n It has 1 dark grey helicopter to its right in view 2. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[96.0, 588.0, 253.0, 687.0]], "image_paths": ["images/img_001006.png", "images/img_000997.png", "images/img_001003.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000945", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'F-16 fighter jet'.\n It has 1 stylized F22 fighter jet with white and red color to its front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[720.0, 625.0, 802.0, 680.0]], "image_paths": ["images/img_000302.png", "images/img_000293.png", "images/img_000304.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000946", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'man wearing brown coat'.\n It has 1 man wearing green hood to its right in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[619.0, 275.0, 659.0, 450.0]], "image_paths": ["images/img_000065.png", "images/img_000058.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000947", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'beige tombstone'.\n It has 1 flat tombstone to its front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[389.0, 358.0, 405.0, 408.0]], "image_paths": ["images/img_000835.png", "images/img_000843.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000948", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'red curved metal bench without armrests'.\n It has 2 soccer balls to its right-front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[280.0, 345.0, 388.0, 395.0]], "image_paths": ["images/img_000589.png", "images/img_000581.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000949", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'tombstone with shoulder top'.\n It has 2 tombstone with pointed tops to its behind in view 3. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[575.0, 300.0, 629.0, 375.0]], "image_paths": ["images/img_001187.png", "images/img_001185.png", "images/img_001178.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000950", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'olive green helicopter'.\n It has 1 cream and black striped hot air balloon to its right-front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[679.0, 392.0, 770.0, 447.0]], "image_paths": ["images/img_001037.png", "images/img_001029.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000951", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'red curved metal bench without armrests'.\n It has 1 beach ball to its left-front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[463.0, 277.0, 544.0, 353.0]], "image_paths": ["images/img_000635.png", "images/img_000628.png", "images/img_000637.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000952", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'woman in pink clothing'.\n It has 1 woman in gray clothing to its left in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[458.0, 362.0, 487.0, 482.0]], "image_paths": ["images/img_001070.png", "images/img_001063.png", "images/img_001073.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000953", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'lamp'.\n It has 1 truck with white and black cab to its left-front in view 2. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[201.0, 588.0, 242.0, 652.0]], "image_paths": ["images/img_000439.png", "images/img_000435.png", "images/img_000441.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000954", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'airplane with yellow livery'.\n It has 1 airplane with light blue livery to its front, and 1 airplane with light blue livery to its behind in view 2. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[509.0, 360.0, 642.0, 420.0]], "image_paths": ["images/img_000217.png", "images/img_000205.png", "images/img_000211.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000955", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'mirage fighter jet'.\n It has 1 pink and maroon striped hot air balloon to its right-front in view 3. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[449.0, 300.0, 549.0, 367.0]], "image_paths": ["images/img_000945.png", "images/img_000942.png", "images/img_000937.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000956", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'airplane with white and orange livery'.\n It has 1 F-16 fighter jet to its left-behind, and 1 airplane with white and orange livery to its behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[490.0, 373.0, 625.0, 433.0]], "image_paths": ["images/img_000289.png", "images/img_000282.png", "images/img_000286.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000957", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'man wearing brown coat'.\n It has 1 man wearing black coat to its left in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[701.0, 342.0, 762.0, 550.0]], "image_paths": ["images/img_000030.png", "images/img_000039.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000958", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'beige upright tombstone with cross'.\n It has 1 tombstone with pointed top to its front in view 2. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[266.0, 323.0, 318.0, 450.0]], "image_paths": ["images/img_001138.png", "images/img_001126.png", "images/img_001132.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000959", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'tombstone without cross'.\n It has 1 white upright tombstone with cross to its right in view 2. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[420.0, 343.0, 458.0, 410.0]], "image_paths": ["images/img_000832.png", "images/img_000821.png", "images/img_000828.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000960", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'airplane with white and orange livery'.\n It has 1 stylized F22 fighter jet with white and red color to its left-front, and 1 airplane with white and orange livery to its front in view 1. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[355.0, 373.0, 490.0, 433.0]], "image_paths": ["images/img_000281.png", "images/img_000283.png", "images/img_000289.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000961", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'green camouflage helicopter'.\n It has 1 olive green helicopter to its behind in view 3. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[693.0, 402.0, 791.0, 472.0]], "image_paths": ["images/img_001042.png", "images/img_001038.png", "images/img_001029.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000962", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'person saluting'.\n It has 1 person saluting to its right, and 1 lifeguard tower to its behind in view 3. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[411.0, 267.0, 444.0, 398.0]], "image_paths": ["images/img_000703.png", "images/img_000702.png", "images/img_000695.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000963", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'beige tombstone'.\n It has 1 black upright tombstone with cross to its right in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[475.0, 330.0, 507.0, 377.0]], "image_paths": ["images/img_000844.png", "images/img_000837.png", "images/img_000845.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000964", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'airplane with red livery'.\n It has 1 airplane with white and green livery and 1 airplane with red livery to its right in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[84.0, 453.0, 230.0, 508.0]], "image_paths": ["images/img_000240.png", "images/img_000234.png", "images/img_000241.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000965", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'man in crouching posture'.\n It has 1 gray colonnade with circular top to its right-front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[380.0, 338.0, 407.0, 393.0]], "image_paths": ["images/img_000821.png", "images/img_000829.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000966", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'woman in white clothing'.\n It has 1 woman wearing black dress to its left in view 3. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[446.0, 342.0, 482.0, 467.0]], "image_paths": ["images/img_001068.png", "images/img_001073.png", "images/img_001060.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000967", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'harrier jumpjet fighter jet'.\n It has 1 black and white striped hot air balloon to its right-front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[665.0, 418.0, 802.0, 477.0]], "image_paths": ["images/img_000910.png", "images/img_000902.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000968", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'woman in pink clothing'.\n It has 1 woman wearing black hat and 1 woman in pink clothing to its front, and 1 woman wearing black hat to its behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[560.0, 390.0, 595.0, 515.0]], "image_paths": ["images/img_001098.png", "images/img_001107.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000969", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'man wearing gray hat'.\n It has 1 bus stop without enclosure to its left-front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[598.0, 260.0, 659.0, 447.0]], "image_paths": ["images/img_000033.png", "images/img_000042.png", "images/img_000037.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000970", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'white flat tombstone with cross'.\n It has 1 white upright tombstone with cross to its right-front in view 2. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[490.0, 255.0, 616.0, 407.0]], "image_paths": ["images/img_001151.png", "images/img_001146.png", "images/img_001154.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000971", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'man wearing a green hoodie and orange shoes'.\n It has 1 tombstone without cross to its right in view 3. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[369.0, 355.0, 405.0, 452.0]], "image_paths": ["images/img_000812.png", "images/img_000811.png", "images/img_000805.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000972", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'black upright tombstone with cross'.\n It has 1 white upright tombstone with cross and 1 black upright tombstone with cross to its left in view 1. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[419.0, 227.0, 455.0, 342.0]], "image_paths": ["images/img_001160.png", "images/img_001168.png", "images/img_001167.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000973", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'worker'.\n It has 1 worker to its behind, and 1 fan to its front in view 1. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[438.0, 273.0, 458.0, 337.0]], "image_paths": ["images/img_000453.png", "images/img_000457.png", "images/img_000461.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000974", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'man wearing red jacket'.\n It has 1 man wearing gray windbreaker to its front in view 3. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[329.0, 298.0, 386.0, 498.0]], "image_paths": ["images/img_000021.png", "images/img_000026.png", "images/img_000017.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000975", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'toy crab'.\n It has 1 blue swim ring to its behind in view 3. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[334.0, 450.0, 389.0, 492.0]], "image_paths": ["images/img_000685.png", "images/img_000687.png", "images/img_000679.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000976", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'green camouflage helicopter'.\n It has 1 green camouflage helicopter to its right, and 1 cream and black striped hot air balloon to its left-front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[679.0, 382.0, 774.0, 450.0]], "image_paths": ["images/img_000990.png", "images/img_000980.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000977", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'man wearing gray hat'.\n It has 1 white supercar to its right-behind in view 3. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[511.0, 197.0, 542.0, 278.0]], "image_paths": ["images/img_000766.png", "images/img_000765.png", "images/img_000757.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000978", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'green camouflage helicopter'.\n It has 1 green camouflage helicopter and 1 desert camouflage helicopter to its right in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[147.0, 482.0, 280.0, 557.0]], "image_paths": ["images/img_001014.png", "images/img_001020.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000979", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'natural wooden bench with armrests'.\n It has 4 baseballs to its right-front, and 1 natural wooden bench with armrests to its right in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[205.0, 548.0, 301.0, 745.0]], "image_paths": ["images/img_000598.png", "images/img_000604.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000980", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'bench made of white painted wood'.\n It has 2 baseballs to its right-behind, and 2 baseballs to its left-behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[445.0, 300.0, 534.0, 392.0]], "image_paths": ["images/img_000606.png", "images/img_000598.png", "images/img_000605.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000981", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'tombstone with pointed top'.\n It has 1 man in suit and tie to its left-front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[626.0, 225.0, 685.0, 323.0]], "image_paths": ["images/img_001125.png", "images/img_001134.png", "images/img_001135.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000982", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'red buttons snowman'.\n It has 1 jacket snowman to its front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[395.0, 318.0, 444.0, 410.0]], "image_paths": ["images/img_000751.png", "images/img_000744.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000983", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'bench made of black metal'.\n It has 1 stone bench without armrests to its left in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[609.0, 295.0, 754.0, 388.0]], "image_paths": ["images/img_000577.png", "images/img_000585.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000984", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'man wearing orange coat'.\n It has 1 taxi to its right-behind in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[441.0, 197.0, 472.0, 298.0]], "image_paths": ["images/img_000788.png", "images/img_000796.png", "images/img_000797.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000985", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'woman in gray clothing'.\n It has 1 food stall with red and white striped roof to its left-front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[396.0, 375.0, 440.0, 508.0]], "image_paths": ["images/img_001119.png", "images/img_001112.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000986", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'green camouflage helicopter'.\n It has 1 green camouflage helicopter to its right, and 1 red and black striped hot air balloon to its right-front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[686.0, 403.0, 799.0, 463.0]], "image_paths": ["images/img_001024.png", "images/img_001014.png", "images/img_001027.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000987", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'woman in pink clothing'.\n It has 1 woman wearing black dress to its right in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[406.0, 378.0, 438.0, 507.0]], "image_paths": ["images/img_001112.png", "images/img_001118.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000988", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'man wearing santa costume'.\n It has 1 man wearing santa costume to its front, and 1 black hat snowman to its right-front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[562.0, 182.0, 601.0, 293.0]], "image_paths": ["images/img_000727.png", "images/img_000736.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000989", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'harrier jumpjet fighter jet'.\n It has 1 harrier jumpjet fighter jet to its right-front, and 1 white and green striped hot air balloon to its left-behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[547.0, 348.0, 636.0, 408.0]], "image_paths": ["images/img_000945.png", "images/img_000937.png", "images/img_000938.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000990", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'olive green helicopter'.\n It has 2 beige sand helicopters to its right, and 1 beige sand helicopter to its left in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[260.0, 397.0, 321.0, 452.0]], "image_paths": ["images/img_000972.png", "images/img_000965.png", "images/img_000975.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000991", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'stylized F22 fighter jet with grey and cyan color'.\n It has 1 stylized F22 fighter jet with grey and cyan color to its right, and 1 F-22 fighter jet with gray color to its right-front in view 1. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[214.0, 368.0, 326.0, 463.0]], "image_paths": ["images/img_000953.png", "images/img_000959.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000992", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'airplane with red livery'.\n It has 1 airplane with red livery and 1 airplane with white and green livery to its front in view 3. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[693.0, 480.0, 861.0, 560.0]], "image_paths": ["images/img_000198.png", "images/img_000202.png", "images/img_000190.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000993", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'man wearing santa costume'.\n It has 1 bus stop with advertisement poster to its right-front in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[319.0, 300.0, 376.0, 497.0]], "image_paths": ["images/img_000067.png", "images/img_000061.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000994", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'swim ring'.\n It has 1 person to its right-behind in view 3. Then return its bounding box in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[544.0, 412.0, 591.0, 460.0]], "image_paths": ["images/img_000670.png", "images/img_000668.png", "images/img_000660.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000995", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'fan'.\n It has 1 truck with white and black cab to its right-behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[579.0, 283.0, 613.0, 333.0]], "image_paths": ["images/img_000461.png", "images/img_000453.png", "images/img_000457.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000996", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'airplane with yellow livery'.\n It has 1 airplane with white and green livery to its right in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[367.0, 378.0, 472.0, 423.0]], "image_paths": ["images/img_000212.png", "images/img_000206.png", "images/img_000217.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000997", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, locate the 'airplane with red livery'.\n It has 1 mirage fighter jet to its left-behind in view 3. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[505.0, 357.0, 619.0, 437.0]], "image_paths": ["images/img_000334.png", "images/img_000329.png", "images/img_000326.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000998", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'worker'.\n It has 1 worker to its behind, and 1 white stack of sandbags to its left-behind in view 2. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[224.0, 368.0, 251.0, 430.0]], "image_paths": ["images/img_000458.png", "images/img_000453.png", "images/img_000460.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_000999", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, locate the 'person saluting'.\n It has 1 person saluting and 1 person with black hat to its front in view 1. Then return its bounding box in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[261.0, 372.0, 314.0, 502.0]], "image_paths": ["images/img_000695.png", "images/img_000704.png", "images/img_000702.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001000", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, locate the 'F-16 fighter jet'.\n It has 1 stylized F22 fighter jet with camoflague color to its behind in view 2. Then return its bounding box in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[443.0, 333.0, 535.0, 377.0]], "image_paths": ["images/img_000957.png", "images/img_000950.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001001", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'bench made of black metal' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[427.0, 225.0, 546.0, 372.0], [696.0, 390.0, 840.0, 557.0]], "image_paths": ["images/img_000639.png", "images/img_000633.png", "images/img_000636.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001002", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'person saluting' that are visible in view 2 but not in view 3. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[632.0, 347.0, 685.0, 493.0]], "image_paths": ["images/img_000682.png", "images/img_000685.png", "images/img_000679.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001003", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'black street lamp' that are visible in view 3 but not in view 1. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[274.0, 485.0, 319.0, 592.0]], "image_paths": ["images/img_000159.png", "images/img_000166.png", "images/img_000165.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001004", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'airplane with yellow livery' that are visible in view 2 but not in view 1. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[180.0, 405.0, 338.0, 473.0], [39.0, 557.0, 269.0, 670.0], [486.0, 557.0, 704.0, 672.0], [704.0, 557.0, 922.0, 672.0]], "image_paths": ["images/img_000250.png", "images/img_000258.png", "images/img_000254.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001005", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'dark grey helicopter' that are visible in view 2 but not in view 3. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[375.0, 347.0, 470.0, 397.0], [522.0, 588.0, 679.0, 687.0]], "image_paths": ["images/img_001006.png", "images/img_001007.png", "images/img_000997.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001006", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'trolley' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[138.0, 597.0, 184.0, 683.0], [159.0, 542.0, 201.0, 620.0], [235.0, 350.0, 266.0, 388.0]], "image_paths": ["images/img_000455.png", "images/img_000447.png", "images/img_000457.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001007", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'stylized F22 fighter jet with camoflague color' that are visible in view 3 but not in view 2. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[224.0, 455.0, 306.0, 517.0]], "image_paths": ["images/img_000273.png", "images/img_000264.png", "images/img_000271.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001008", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'red shipping container' that are visible in view 1 but not in view 3. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[256.0, 345.0, 314.0, 447.0]], "image_paths": ["images/img_000082.png", "images/img_000081.png", "images/img_000075.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001009", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'woman in gray clothing' that are visible in view 1 but not in view 3. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[507.0, 448.0, 555.0, 590.0], [374.0, 445.0, 421.0, 607.0]], "image_paths": ["images/img_001069.png", "images/img_001073.png", "images/img_001059.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001010", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'blue shipping container' that are visible in view 2 but not in view 3. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[675.0, 433.0, 792.0, 515.0], [479.0, 283.0, 517.0, 380.0]], "image_paths": ["images/img_000102.png", "images/img_000106.png", "images/img_000101.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001011", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'harrier jumpjet fighter jet' that are visible in view 3 but not in view 2. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[286.0, 608.0, 481.0, 698.0], [657.0, 415.0, 795.0, 473.0], [723.0, 608.0, 916.0, 698.0]], "image_paths": ["images/img_000944.png", "images/img_000930.png", "images/img_000938.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001012", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'tombstone with pointed top' that are visible in view 3 but not in view 1. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[186.0, 390.0, 263.0, 513.0]], "image_paths": ["images/img_001126.png", "images/img_001132.png", "images/img_001135.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001013", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'trellis' that are visible in view 2 but not in view 3. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[361.0, 222.0, 456.0, 382.0], [142.0, 363.0, 278.0, 547.0]], "image_paths": ["images/img_000175.png", "images/img_000173.png", "images/img_000167.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001014", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'drone with white oval body shape' that are visible in view 2 but not in view 3. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[740.0, 405.0, 767.0, 427.0], [569.0, 655.0, 609.0, 702.0], [417.0, 653.0, 481.0, 693.0], [631.0, 365.0, 670.0, 383.0]], "image_paths": ["images/img_000349.png", "images/img_000347.png", "images/img_000340.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001015", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'beach ball' that are visible in view 3 but not in view 2. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[610.0, 398.0, 644.0, 443.0]], "image_paths": ["images/img_000718.png", "images/img_000708.png", "images/img_000714.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001016", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'red shipping container' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[605.0, 583.0, 680.0, 642.0]], "image_paths": ["images/img_000550.png", "images/img_000546.png", "images/img_000557.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001017", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'airplane with white and orange livery' that are visible in view 1 but not in view 3. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[641.0, 373.0, 765.0, 418.0]], "image_paths": ["images/img_000226.png", "images/img_000229.png", "images/img_000220.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001018", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'drone with smooth black curved body' that are visible in view 2 but not in view 3. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[552.0, 658.0, 675.0, 748.0], [99.0, 525.0, 204.0, 583.0], [334.0, 668.0, 461.0, 760.0]], "image_paths": ["images/img_000384.png", "images/img_000385.png", "images/img_000379.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001019", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'blue shipping container' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[530.0, 278.0, 624.0, 378.0], [330.0, 640.0, 469.0, 830.0]], "image_paths": ["images/img_000131.png", "images/img_000125.png", "images/img_000127.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001020", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'white upright tombstone with cross' that are visible in view 2 but not in view 1. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[573.0, 585.0, 630.0, 732.0]], "image_paths": ["images/img_001195.png", "images/img_001202.png", "images/img_001203.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001021", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'woman in gray clothing' that are visible in view 3 but not in view 1. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[574.0, 448.0, 621.0, 590.0], [507.0, 448.0, 555.0, 590.0]], "image_paths": ["images/img_001063.png", "images/img_001067.png", "images/img_001069.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001022", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'mirage fighter jet' that are visible in view 3 but not in view 2. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[216.0, 367.0, 347.0, 453.0]], "image_paths": ["images/img_000941.png", "images/img_000937.png", "images/img_000945.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001023", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'worker' that are visible in view 2 but not in view 3. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[407.0, 713.0, 441.0, 808.0]], "image_paths": ["images/img_000441.png", "images/img_000444.png", "images/img_000436.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001024", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'worker wearing red hat and white shirt' that are visible in view 3 but not in view 1. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[703.0, 582.0, 733.0, 660.0]], "image_paths": ["images/img_000466.png", "images/img_000473.png", "images/img_000475.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001025", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'person with tattoos' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[302.0, 388.0, 347.0, 527.0]], "image_paths": ["images/img_000654.png", "images/img_000646.png", "images/img_000655.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001026", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'blue trellis' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[733.0, 537.0, 790.0, 790.0]], "image_paths": ["images/img_000139.png", "images/img_000134.png", "images/img_000144.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001027", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'airplane with yellow livery' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[233.0, 377.0, 369.0, 467.0], [590.0, 512.0, 755.0, 658.0]], "image_paths": ["images/img_000231.png", "images/img_000222.png", "images/img_000230.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001028", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'blue shipping container' that are visible in view 3 but not in view 2. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[721.0, 640.0, 874.0, 745.0], [206.0, 437.0, 325.0, 518.0]], "image_paths": ["images/img_000108.png", "images/img_000098.png", "images/img_000104.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001029", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'man holding box' that are visible in view 2 but not in view 3. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[366.0, 383.0, 394.0, 463.0]], "image_paths": ["images/img_000846.png", "images/img_000845.png", "images/img_000835.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001030", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'woman in white clothing' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[472.0, 488.0, 519.0, 630.0], [434.0, 455.0, 477.0, 598.0]], "image_paths": ["images/img_001056.png", "images/img_001049.png", "images/img_001051.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001031", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'airplane with light blue livery' that are visible in view 3 but not in view 1. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[624.0, 363.0, 759.0, 467.0]], "image_paths": ["images/img_000307.png", "images/img_000318.png", "images/img_000312.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001032", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'concrete mixer' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[838.0, 417.0, 892.0, 482.0], [802.0, 395.0, 849.0, 463.0]], "image_paths": ["images/img_000441.png", "images/img_000435.png", "images/img_000443.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001033", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'man wearing green hood' that are visible in view 1 but not in view 3. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[391.0, 388.0, 458.0, 653.0]], "image_paths": ["images/img_000069.png", "images/img_000072.png", "images/img_000060.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001034", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'tombstone with shoulder top' that are visible in view 2 but not in view 1. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[346.0, 620.0, 422.0, 725.0], [660.0, 522.0, 733.0, 620.0]], "image_paths": ["images/img_001174.png", "images/img_001183.png", "images/img_001185.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001035", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'red buttons snowman' that are visible in view 1 but not in view 3. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[578.0, 322.0, 644.0, 410.0]], "image_paths": ["images/img_000799.png", "images/img_000798.png", "images/img_000792.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001036", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'green camouflage helicopter' that are visible in view 3 but not in view 1. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[729.0, 470.0, 846.0, 570.0], [156.0, 470.0, 273.0, 570.0]], "image_paths": ["images/img_001035.png", "images/img_001037.png", "images/img_001042.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001037", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'airplane with white and green livery' that are visible in view 1 but not in view 3. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[703.0, 558.0, 921.0, 673.0]], "image_paths": ["images/img_000200.png", "images/img_000197.png", "images/img_000193.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001038", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'airplane with yellow livery' that are visible in view 3 but not in view 1. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[510.0, 550.0, 669.0, 632.0]], "image_paths": ["images/img_000294.png", "images/img_000298.png", "images/img_000302.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001039", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'stone bench without armrests' that are visible in view 3 but not in view 2. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[427.0, 630.0, 541.0, 790.0]], "image_paths": ["images/img_000584.png", "images/img_000582.png", "images/img_000590.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001040", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'flat tombstone' that are visible in view 1 but not in view 3. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[522.0, 557.0, 569.0, 655.0]], "image_paths": ["images/img_000847.png", "images/img_000843.png", "images/img_000841.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001041", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'man wearing red jacket' that are visible in view 2 but not in view 3. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[323.0, 385.0, 385.0, 610.0]], "image_paths": ["images/img_000009.png", "images/img_000011.png", "images/img_000004.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001042", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'white shipping container' that are visible in view 2 but not in view 1. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[396.0, 272.0, 492.0, 370.0]], "image_paths": ["images/img_000122.png", "images/img_000129.png", "images/img_000131.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001043", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'olive green helicopter' that are visible in view 3 but not in view 2. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[104.0, 442.0, 163.0, 505.0], [250.0, 392.0, 311.0, 447.0]], "image_paths": ["images/img_001041.png", "images/img_001029.png", "images/img_001037.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001044", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'stone steps' that are visible in view 1 but not in view 3. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[371.0, 533.0, 649.0, 873.0]], "image_paths": ["images/img_000409.png", "images/img_000415.png", "images/img_000402.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001045", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'drone with white oval body shape' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[887.0, 673.0, 936.0, 723.0], [746.0, 417.0, 791.0, 433.0], [240.0, 455.0, 289.0, 478.0]], "image_paths": ["images/img_000397.png", "images/img_000390.png", "images/img_000398.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001046", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'woman wearing underwear' that are visible in view 2 but not in view 3. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[444.0, 437.0, 490.0, 602.0]], "image_paths": ["images/img_001084.png", "images/img_001087.png", "images/img_001081.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001047", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'drone with white oval body shape' that are visible in view 2 but not in view 1. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[795.0, 505.0, 844.0, 532.0]], "image_paths": ["images/img_000378.png", "images/img_000383.png", "images/img_000381.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001048", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'green shipping container' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[772.0, 418.0, 876.0, 547.0]], "image_paths": ["images/img_000093.png", "images/img_000089.png", "images/img_000091.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001049", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'stone steps' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[600.0, 273.0, 761.0, 438.0]], "image_paths": ["images/img_000412.png", "images/img_000405.png", "images/img_000410.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001050", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'airplane with red livery' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[156.0, 432.0, 318.0, 565.0], [206.0, 380.0, 344.0, 487.0]], "image_paths": ["images/img_000202.png", "images/img_000194.png", "images/img_000201.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001051", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'beige tombstone' that are visible in view 3 but not in view 1. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[781.0, 398.0, 844.0, 538.0]], "image_paths": ["images/img_001145.png", "images/img_001153.png", "images/img_001152.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001052", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'harrier jumpjet fighter jet' that are visible in view 3 but not in view 1. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[464.0, 572.0, 531.0, 640.0], [815.0, 478.0, 870.0, 528.0]], "image_paths": ["images/img_000320.png", "images/img_000332.png", "images/img_000327.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001053", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'trolley' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[718.0, 368.0, 774.0, 407.0], [769.0, 493.0, 841.0, 538.0]], "image_paths": ["images/img_000408.png", "images/img_000402.png", "images/img_000409.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001054", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'man wearing black coat' that are visible in view 3 but not in view 1. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[681.0, 325.0, 738.0, 545.0]], "image_paths": ["images/img_000045.png", "images/img_000050.png", "images/img_000051.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001055", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'striped scarf snowman' that are visible in view 2 but not in view 3. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[566.0, 533.0, 630.0, 630.0]], "image_paths": ["images/img_000736.png", "images/img_000730.png", "images/img_000723.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001056", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'woman wearing black hat' that are visible in view 2 but not in view 1. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[611.0, 405.0, 642.0, 538.0]], "image_paths": ["images/img_001099.png", "images/img_001107.png", "images/img_001105.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001057", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'airplane with white and green livery' that are visible in view 1 but not in view 3. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[59.0, 565.0, 265.0, 683.0]], "image_paths": ["images/img_000256.png", "images/img_000257.png", "images/img_000250.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001058", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'tennis ball' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[151.0, 608.0, 164.0, 625.0], [840.0, 450.0, 851.0, 463.0]], "image_paths": ["images/img_000607.png", "images/img_000598.png", "images/img_000604.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001059", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'basketball' that are visible in view 3 but not in view 1. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[176.0, 537.0, 196.0, 563.0], [171.0, 553.0, 193.0, 580.0], [181.0, 522.0, 201.0, 548.0], [795.0, 555.0, 816.0, 582.0], [744.0, 402.0, 761.0, 425.0]], "image_paths": ["images/img_000610.png", "images/img_000621.png", "images/img_000617.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001060", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'red shipping container' that are visible in view 2 but not in view 1. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[704.0, 642.0, 845.0, 745.0], [593.0, 615.0, 661.0, 780.0], [672.0, 345.0, 730.0, 450.0]], "image_paths": ["images/img_000113.png", "images/img_000118.png", "images/img_000116.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001061", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'woman wearing scarf accessory' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[611.0, 405.0, 642.0, 538.0]], "image_paths": ["images/img_001107.png", "images/img_001099.png", "images/img_001105.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001062", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'worker' that are visible in view 3 but not in view 2. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[406.0, 705.0, 445.0, 803.0]], "image_paths": ["images/img_000459.png", "images/img_000453.png", "images/img_000461.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001063", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'striped scarf snowman' that are visible in view 2 but not in view 3. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[290.0, 440.0, 364.0, 538.0]], "image_paths": ["images/img_000733.png", "images/img_000736.png", "images/img_000726.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001064", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'beach ball' that are visible in view 3 but not in view 2. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[792.0, 417.0, 807.0, 437.0], [198.0, 613.0, 216.0, 637.0], [639.0, 333.0, 651.0, 350.0], [379.0, 818.0, 401.0, 848.0]], "image_paths": ["images/img_000639.png", "images/img_000631.png", "images/img_000637.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001065", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'dark grey helicopter' that are visible in view 2 but not in view 3. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[215.0, 393.0, 310.0, 470.0]], "image_paths": ["images/img_000975.png", "images/img_000977.png", "images/img_000969.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001066", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'black flat tombstone with cross' that are visible in view 2 but not in view 1. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[600.0, 385.0, 731.0, 475.0]], "image_paths": ["images/img_001159.png", "images/img_001165.png", "images/img_001166.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001067", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'red shipping container' that are visible in view 3 but not in view 1. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[556.0, 373.0, 584.0, 438.0], [649.0, 548.0, 690.0, 637.0]], "image_paths": ["images/img_000504.png", "images/img_000510.png", "images/img_000511.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001068", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'green street lamp' that are visible in view 1 but not in view 3. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[568.0, 607.0, 621.0, 723.0], [425.0, 215.0, 464.0, 325.0]], "image_paths": ["images/img_000150.png", "images/img_000152.png", "images/img_000145.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001069", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'blue shipping container' that are visible in view 3 but not in view 1. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[331.0, 423.0, 395.0, 500.0], [271.0, 460.0, 336.0, 542.0]], "image_paths": ["images/img_000545.png", "images/img_000556.png", "images/img_000551.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001070", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'man wearing gray windbreaker' that are visible in view 3 but not in view 1. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[621.0, 305.0, 666.0, 505.0], [632.0, 345.0, 682.0, 558.0], [541.0, 442.0, 634.0, 675.0]], "image_paths": ["images/img_000020.png", "images/img_000024.png", "images/img_000027.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001071", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'flat tombstone' that are visible in view 2 but not in view 3. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[334.0, 570.0, 421.0, 628.0]], "image_paths": ["images/img_000846.png", "images/img_000845.png", "images/img_000836.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001072", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'surfboard' that are visible in view 2 but not in view 1. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[517.0, 497.0, 581.0, 725.0], [664.0, 293.0, 711.0, 540.0], [255.0, 328.0, 307.0, 577.0]], "image_paths": ["images/img_000664.png", "images/img_000672.png", "images/img_000665.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001073", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'drone with smooth black curved body' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[412.0, 330.0, 480.0, 365.0], [256.0, 587.0, 367.0, 668.0], [529.0, 332.0, 596.0, 367.0]], "image_paths": ["images/img_000383.png", "images/img_000376.png", "images/img_000381.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001074", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'green camouflage helicopter' that are visible in view 2 but not in view 1. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[679.0, 382.0, 774.0, 450.0], [824.0, 430.0, 930.0, 510.0]], "image_paths": ["images/img_000982.png", "images/img_000990.png", "images/img_000988.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001075", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'man in crouching posture' that are visible in view 3 but not in view 1. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[326.0, 458.0, 354.0, 512.0]], "image_paths": ["images/img_000821.png", "images/img_000834.png", "images/img_000830.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001076", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'snowman with black hat and red gloves' that are visible in view 3 but not in view 2. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[450.0, 612.0, 537.0, 732.0], [274.0, 395.0, 361.0, 512.0]], "image_paths": ["images/img_000751.png", "images/img_000739.png", "images/img_000747.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001077", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'green shipping container' that are visible in view 2 but not in view 3. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[594.0, 480.0, 677.0, 538.0], [265.0, 712.0, 374.0, 788.0], [581.0, 330.0, 650.0, 377.0]], "image_paths": ["images/img_000542.png", "images/img_000540.png", "images/img_000534.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001078", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'airplane with white and orange livery' that are visible in view 3 but not in view 2. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[206.0, 380.0, 344.0, 487.0]], "image_paths": ["images/img_000243.png", "images/img_000237.png", "images/img_000245.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001079", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'trellis' that are visible in view 1 but not in view 3. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[564.0, 543.0, 585.0, 800.0], [686.0, 408.0, 729.0, 635.0]], "image_paths": ["images/img_000176.png", "images/img_000174.png", "images/img_000169.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001080", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'drone with smooth black curved body' that are visible in view 2 but not in view 3. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[421.0, 360.0, 492.0, 393.0], [677.0, 602.0, 800.0, 683.0]], "image_paths": ["images/img_000361.png", "images/img_000360.png", "images/img_000355.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001081", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'blue swim ring' that are visible in view 2 but not in view 1. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[340.0, 542.0, 400.0, 605.0], [465.0, 687.0, 531.0, 763.0]], "image_paths": ["images/img_000693.png", "images/img_000702.png", "images/img_000700.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001082", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'green shipping container' that are visible in view 3 but not in view 1. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[440.0, 543.0, 484.0, 667.0], [689.0, 338.0, 736.0, 423.0]], "image_paths": ["images/img_000488.png", "images/img_000490.png", "images/img_000495.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001083", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'F-22 fighter jet with gray color' that are visible in view 2 but not in view 1. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[226.0, 600.0, 421.0, 720.0]], "image_paths": ["images/img_000902.png", "images/img_000909.png", "images/img_000912.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001084", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'man wearing a green hoodie and orange shoes' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[446.0, 323.0, 477.0, 417.0], [383.0, 472.0, 409.0, 572.0]], "image_paths": ["images/img_000866.png", "images/img_000857.png", "images/img_000864.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001085", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'white upright tombstone with cross' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[370.0, 592.0, 422.0, 725.0]], "image_paths": ["images/img_001203.png", "images/img_001196.png", "images/img_001202.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001086", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'F-16 fighter jet' that are visible in view 3 but not in view 1. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[776.0, 610.0, 864.0, 667.0]], "image_paths": ["images/img_000309.png", "images/img_000319.png", "images/img_000317.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001087", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'F-22 fighter jet with gray color' that are visible in view 3 but not in view 2. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[383.0, 368.0, 481.0, 423.0]], "image_paths": ["images/img_000955.png", "images/img_000946.png", "images/img_000956.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001088", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'worker' that are visible in view 1 but not in view 3. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[657.0, 305.0, 682.0, 370.0], [856.0, 403.0, 887.0, 483.0]], "image_paths": ["images/img_000425.png", "images/img_000429.png", "images/img_000419.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001089", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'drone with smooth black curved body' that are visible in view 2 but not in view 1. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[759.0, 577.0, 853.0, 653.0], [724.0, 515.0, 818.0, 562.0], [228.0, 675.0, 345.0, 748.0]], "image_paths": ["images/img_000363.png", "images/img_000369.png", "images/img_000371.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001090", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'man wearing brown coat' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[646.0, 323.0, 694.0, 515.0], [461.0, 490.0, 525.0, 737.0]], "image_paths": ["images/img_000041.png", "images/img_000035.png", "images/img_000043.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001091", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'green camouflage helicopter' that are visible in view 1 but not in view 3. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[66.0, 430.0, 188.0, 510.0], [676.0, 382.0, 771.0, 448.0]], "image_paths": ["images/img_001023.png", "images/img_001024.png", "images/img_001015.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001092", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'man holding box' that are visible in view 3 but not in view 1. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[364.0, 530.0, 389.0, 620.0]], "image_paths": ["images/img_000838.png", "images/img_000843.png", "images/img_000847.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001093", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'man wearing brown coat' that are visible in view 2 but not in view 3. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[601.0, 267.0, 640.0, 438.0]], "image_paths": ["images/img_000011.png", "images/img_000007.png", "images/img_000002.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001094", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'man wearing gray windbreaker' that are visible in view 2 but not in view 1. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[480.0, 228.0, 516.0, 313.0], [335.0, 227.0, 375.0, 312.0], [378.0, 228.0, 401.0, 317.0]], "image_paths": ["images/img_000771.png", "images/img_000781.png", "images/img_000780.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001095", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'three balls snowman' that are visible in view 2 but not in view 3. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[637.0, 433.0, 695.0, 543.0]], "image_paths": ["images/img_000764.png", "images/img_000763.png", "images/img_000755.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001096", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'trolley' that are visible in view 3 but not in view 1. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[859.0, 483.0, 912.0, 552.0]], "image_paths": ["images/img_000419.png", "images/img_000426.png", "images/img_000427.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001097", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'tombstone with pointed top' that are visible in view 3 but not in view 2. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[671.0, 637.0, 755.0, 752.0]], "image_paths": ["images/img_001137.png", "images/img_001125.png", "images/img_001132.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001098", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'worker' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[102.0, 478.0, 133.0, 555.0], [657.0, 305.0, 682.0, 370.0]], "image_paths": ["images/img_000425.png", "images/img_000417.png", "images/img_000426.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001099", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'airplane with yellow livery' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[315.0, 535.0, 515.0, 637.0]], "image_paths": ["images/img_000299.png", "images/img_000291.png", "images/img_000301.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001100", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'red shipping container' that are visible in view 3 but not in view 2. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[399.0, 542.0, 434.0, 628.0], [564.0, 542.0, 599.0, 628.0]], "image_paths": ["images/img_000505.png", "images/img_000500.png", "images/img_000507.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001101", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'toy crab' that are visible in view 1 but not in view 3. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[307.0, 443.0, 359.0, 490.0]], "image_paths": ["images/img_000720.png", "images/img_000716.png", "images/img_000712.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001102", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'drone with smooth black curved body' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[196.0, 557.0, 309.0, 627.0], [677.0, 602.0, 800.0, 683.0]], "image_paths": ["images/img_000360.png", "images/img_000353.png", "images/img_000362.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001103", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'green street lamp' that are visible in view 2 but not in view 1. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[372.0, 592.0, 414.0, 705.0], [596.0, 592.0, 637.0, 705.0]], "image_paths": ["images/img_000179.png", "images/img_000186.png", "images/img_000185.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001104", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'drone with angular black body' that are visible in view 2 but not in view 3. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[562.0, 630.0, 752.0, 777.0]], "image_paths": ["images/img_000359.png", "images/img_000360.png", "images/img_000355.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001105", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'blue swim ring' that are visible in view 2 but not in view 3. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[346.0, 555.0, 406.0, 620.0], [429.0, 648.0, 492.0, 722.0]], "image_paths": ["images/img_000702.png", "images/img_000698.png", "images/img_000689.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001106", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'black street lamp' that are visible in view 3 but not in view 1. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[590.0, 577.0, 614.0, 692.0]], "image_paths": ["images/img_000158.png", "images/img_000164.png", "images/img_000163.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001107", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'airplane with light blue livery' that are visible in view 3 but not in view 1. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[631.0, 377.0, 767.0, 467.0]], "image_paths": ["images/img_000237.png", "images/img_000245.png", "images/img_000244.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001108", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'baseball' that are visible in view 3 but not in view 2. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[699.0, 710.0, 714.0, 730.0], [864.0, 527.0, 877.0, 543.0]], "image_paths": ["images/img_000570.png", "images/img_000563.png", "images/img_000571.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001109", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'red gloves snowman' that are visible in view 3 but not in view 1. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[470.0, 222.0, 541.0, 332.0]], "image_paths": ["images/img_000776.png", "images/img_000779.png", "images/img_000782.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001110", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'green shipping container' that are visible in view 3 but not in view 2. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[431.0, 545.0, 470.0, 652.0]], "image_paths": ["images/img_000525.png", "images/img_000519.png", "images/img_000526.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001111", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'bench made of white painted wood' that are visible in view 1 but not in view 3. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[440.0, 602.0, 566.0, 697.0]], "image_paths": ["images/img_000606.png", "images/img_000607.png", "images/img_000597.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001112", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'trellis' that are visible in view 2 but not in view 3. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[623.0, 350.0, 762.0, 448.0]], "image_paths": ["images/img_000172.png", "images/img_000174.png", "images/img_000168.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001113", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'red shipping container' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[616.0, 625.0, 703.0, 692.0]], "image_paths": ["images/img_000540.png", "images/img_000533.png", "images/img_000539.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001114", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'olive green helicopter' that are visible in view 2 but not in view 3. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[820.0, 582.0, 882.0, 708.0], [364.0, 582.0, 410.0, 708.0]], "image_paths": ["images/img_000991.png", "images/img_000989.png", "images/img_000980.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001115", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'blue street lamp' that are visible in view 3 but not in view 2. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[366.0, 592.0, 405.0, 708.0]], "image_paths": ["images/img_000181.png", "images/img_000179.png", "images/img_000184.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001116", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'airplane with white and green livery' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[141.0, 455.0, 294.0, 577.0], [62.0, 412.0, 206.0, 515.0]], "image_paths": ["images/img_000229.png", "images/img_000222.png", "images/img_000230.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001117", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'tombstone with pointed top' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[682.0, 588.0, 765.0, 703.0], [642.0, 333.0, 706.0, 437.0]], "image_paths": ["images/img_001184.png", "images/img_001178.png", "images/img_001181.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001118", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'woman in gray clothing' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[559.0, 370.0, 606.0, 512.0]], "image_paths": ["images/img_001088.png", "images/img_001080.png", "images/img_001083.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001119", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'man wearing black coat' that are visible in view 2 but not in view 3. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[339.0, 295.0, 391.0, 497.0]], "image_paths": ["images/img_000007.png", "images/img_000014.png", "images/img_000004.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001120", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'green camouflage helicopter' that are visible in view 3 but not in view 2. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[311.0, 577.0, 455.0, 713.0], [75.0, 577.0, 219.0, 713.0]], "image_paths": ["images/img_000989.png", "images/img_000983.png", "images/img_000993.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001121", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'snowman with black hat and black buttons' that are visible in view 1 but not in view 3. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[450.0, 612.0, 537.0, 732.0]], "image_paths": ["images/img_000747.png", "images/img_000751.png", "images/img_000740.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001122", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'man wearing a green hoodie and orange shoes' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[375.0, 445.0, 401.0, 545.0], [551.0, 508.0, 589.0, 602.0]], "image_paths": ["images/img_000864.png", "images/img_000857.png", "images/img_000862.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001123", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'man wearing santa costume' that are visible in view 2 but not in view 3. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[389.0, 663.0, 432.0, 763.0], [414.0, 717.0, 458.0, 798.0]], "image_paths": ["images/img_000801.png", "images/img_000798.png", "images/img_000791.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001124", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'man in crouching posture' that are visible in view 3 but not in view 1. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[492.0, 312.0, 517.0, 363.0]], "image_paths": ["images/img_000822.png", "images/img_000834.png", "images/img_000832.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001125", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'green street lamp' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[561.0, 623.0, 616.0, 742.0]], "image_paths": ["images/img_000154.png", "images/img_000147.png", "images/img_000150.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001126", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'stylized F22 fighter jet with camoflague color' that are visible in view 2 but not in view 1. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[259.0, 525.0, 356.0, 600.0]], "image_paths": ["images/img_000264.png", "images/img_000270.png", "images/img_000273.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001127", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'stylized F22 fighter jet with white and red color' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[455.0, 587.0, 559.0, 655.0]], "image_paths": ["images/img_000287.png", "images/img_000280.png", "images/img_000290.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001128", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'concrete mixer' that are visible in view 2 but not in view 1. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[525.0, 833.0, 589.0, 927.0], [568.0, 775.0, 631.0, 885.0]], "image_paths": ["images/img_000435.png", "images/img_000443.png", "images/img_000438.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001129", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'olive green helicopter' that are visible in view 1 but not in view 3. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[237.0, 357.0, 331.0, 397.0]], "image_paths": ["images/img_000975.png", "images/img_000976.png", "images/img_000966.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001130", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'bench made of natural wood' that are visible in view 1 but not in view 3. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[410.0, 618.0, 580.0, 847.0], [164.0, 378.0, 282.0, 542.0]], "image_paths": ["images/img_000603.png", "images/img_000604.png", "images/img_000594.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001131", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'drone with angular white body and gray propellers' that are visible in view 3 but not in view 2. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[835.0, 623.0, 919.0, 690.0]], "image_paths": ["images/img_000395.png", "images/img_000391.png", "images/img_000399.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001132", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'mirage fighter jet' that are visible in view 3 but not in view 2. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[338.0, 528.0, 450.0, 718.0]], "image_paths": ["images/img_000891.png", "images/img_000889.png", "images/img_000897.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001133", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'man in crouching posture' that are visible in view 3 but not in view 2. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[395.0, 352.0, 421.0, 403.0]], "image_paths": ["images/img_000833.png", "images/img_000822.png", "images/img_000831.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001134", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'surfboard' that are visible in view 3 but not in view 1. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[619.0, 415.0, 681.0, 652.0], [276.0, 498.0, 357.0, 753.0]], "image_paths": ["images/img_000663.png", "images/img_000665.png", "images/img_000671.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001135", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'woman wearing underwear' that are visible in view 1 but not in view 3. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[586.0, 390.0, 624.0, 550.0]], "image_paths": ["images/img_001083.png", "images/img_001090.png", "images/img_001075.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001136", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'airplane with white and orange livery' that are visible in view 2 but not in view 3. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[357.0, 348.0, 472.0, 428.0]], "image_paths": ["images/img_000241.png", "images/img_000246.png", "images/img_000238.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001137", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'beige tombstone' that are visible in view 3 but not in view 1. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[494.0, 170.0, 535.0, 298.0]], "image_paths": ["images/img_001143.png", "images/img_001151.png", "images/img_001152.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001138", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'airplane with white and green livery' that are visible in view 2 but not in view 1. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[86.0, 503.0, 280.0, 678.0], [510.0, 505.0, 703.0, 680.0]], "image_paths": ["images/img_000210.png", "images/img_000217.png", "images/img_000211.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001139", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'airplane with white and orange livery' that are visible in view 1 but not in view 3. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[639.0, 492.0, 780.0, 555.0]], "image_paths": ["images/img_000302.png", "images/img_000298.png", "images/img_000294.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001140", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'bench made of black metal' that are visible in view 2 but not in view 3. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[614.0, 282.0, 748.0, 397.0], [426.0, 550.0, 606.0, 717.0]], "image_paths": ["images/img_000566.png", "images/img_000568.png", "images/img_000562.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001141", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'black street lamp' that are visible in view 1 but not in view 3. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[677.0, 502.0, 718.0, 615.0]], "image_paths": ["images/img_000142.png", "images/img_000141.png", "images/img_000136.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001142", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'surfboard' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[625.0, 370.0, 699.0, 620.0]], "image_paths": ["images/img_000654.png", "images/img_000645.png", "images/img_000655.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001143", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'stylized F22 fighter jet with white and red color' that are visible in view 3 but not in view 1. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[521.0, 552.0, 715.0, 722.0]], "image_paths": ["images/img_000904.png", "images/img_000906.png", "images/img_000910.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001144", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'dark grey helicopter' that are visible in view 3 but not in view 1. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[164.0, 483.0, 276.0, 573.0], [689.0, 407.0, 786.0, 477.0]], "image_paths": ["images/img_000998.png", "images/img_001009.png", "images/img_001005.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001145", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'stylized F22 fighter jet with white and red color' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[576.0, 530.0, 755.0, 693.0]], "image_paths": ["images/img_000929.png", "images/img_000920.png", "images/img_000923.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001146", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'woman wearing black dress' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[361.0, 390.0, 385.0, 533.0]], "image_paths": ["images/img_001118.png", "images/img_001109.png", "images/img_001122.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001147", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'airplane with red livery' that are visible in view 3 but not in view 2. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[289.0, 498.0, 444.0, 633.0]], "image_paths": ["images/img_000328.png", "images/img_000325.png", "images/img_000334.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001148", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'forklift' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[203.0, 460.0, 237.0, 512.0]], "image_paths": ["images/img_000510.png", "images/img_000502.png", "images/img_000507.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001149", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'tennis ball' that are visible in view 3 but not in view 1. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[196.0, 608.0, 210.0, 625.0]], "image_paths": ["images/img_000582.png", "images/img_000585.png", "images/img_000590.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001150", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'green shipping container' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[698.0, 468.0, 833.0, 612.0], [601.0, 317.0, 695.0, 423.0], [310.0, 618.0, 458.0, 800.0]], "image_paths": ["images/img_000081.png", "images/img_000075.png", "images/img_000080.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001151", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'woman in white clothing' that are visible in view 3 but not in view 2. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[580.0, 415.0, 611.0, 550.0]], "image_paths": ["images/img_001067.png", "images/img_001063.png", "images/img_001071.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001152", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'man wearing black coat' that are visible in view 2 but not in view 3. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[603.0, 262.0, 651.0, 462.0]], "image_paths": ["images/img_000053.png", "images/img_000054.png", "images/img_000048.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001153", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'beige tombstone' that are visible in view 2 but not in view 3. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[205.0, 367.0, 264.0, 510.0]], "image_paths": ["images/img_001153.png", "images/img_001150.png", "images/img_001142.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001154", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'white shipping container' that are visible in view 1 but not in view 3. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[175.0, 485.0, 294.0, 635.0]], "image_paths": ["images/img_000115.png", "images/img_000121.png", "images/img_000110.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001155", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'blue shipping container' that are visible in view 1 but not in view 3. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[156.0, 607.0, 253.0, 778.0], [675.0, 433.0, 792.0, 515.0]], "image_paths": ["images/img_000106.png", "images/img_000105.png", "images/img_000100.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001156", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'beach ball' that are visible in view 2 but not in view 1. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[444.0, 317.0, 477.0, 362.0]], "image_paths": ["images/img_000676.png", "images/img_000682.png", "images/img_000683.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001157", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'white shipping container' that are visible in view 3 but not in view 1. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[599.0, 497.0, 701.0, 610.0], [458.0, 612.0, 565.0, 748.0], [326.0, 390.0, 406.0, 483.0]], "image_paths": ["images/img_000481.png", "images/img_000491.png", "images/img_000490.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001158", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'F-22 fighter jet with gray color' that are visible in view 1 but not in view 3. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[659.0, 368.0, 759.0, 423.0], [715.0, 500.0, 851.0, 600.0]], "image_paths": ["images/img_000956.png", "images/img_000954.png", "images/img_000949.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001159", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'man wearing a green hoodie and orange shoes' that are visible in view 3 but not in view 2. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[596.0, 438.0, 621.0, 535.0]], "image_paths": ["images/img_000864.png", "images/img_000857.png", "images/img_000863.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001160", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'black street lamp' that are visible in view 3 but not in view 2. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[686.0, 363.0, 728.0, 470.0]], "image_paths": ["images/img_000137.png", "images/img_000136.png", "images/img_000141.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001161", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'drone with angular black body' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[355.0, 638.0, 500.0, 752.0]], "image_paths": ["images/img_000369.png", "images/img_000365.png", "images/img_000368.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001162", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'airplane with white and green livery' that are visible in view 3 but not in view 1. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[642.0, 375.0, 766.0, 420.0], [371.0, 373.0, 495.0, 418.0]], "image_paths": ["images/img_000208.png", "images/img_000215.png", "images/img_000213.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001163", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'baseball' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[791.0, 523.0, 805.0, 540.0], [733.0, 348.0, 743.0, 362.0]], "image_paths": ["images/img_000570.png", "images/img_000563.png", "images/img_000569.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001164", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'bench made of blue painted metal' that are visible in view 3 but not in view 1. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[564.0, 285.0, 724.0, 390.0]], "image_paths": ["images/img_000612.png", "images/img_000617.png", "images/img_000619.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001165", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'green shipping container' that are visible in view 3 but not in view 1. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[431.0, 545.0, 470.0, 652.0]], "image_paths": ["images/img_000518.png", "images/img_000523.png", "images/img_000526.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001166", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'man in crouching posture' that are visible in view 2 but not in view 1. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[615.0, 355.0, 640.0, 405.0]], "image_paths": ["images/img_000824.png", "images/img_000831.png", "images/img_000833.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001167", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'beach ball' that are visible in view 1 but not in view 3. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[613.0, 607.0, 651.0, 658.0]], "image_paths": ["images/img_000652.png", "images/img_000649.png", "images/img_000646.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001168", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'F-16 fighter jet' that are visible in view 1 but not in view 3. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[424.0, 633.0, 606.0, 807.0]], "image_paths": ["images/img_000961.png", "images/img_000954.png", "images/img_000952.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001169", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'man wearing red jacket' that are visible in view 3 but not in view 2. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[282.0, 445.0, 352.0, 672.0]], "image_paths": ["images/img_000067.png", "images/img_000063.png", "images/img_000071.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001170", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'green shipping container' that are visible in view 3 but not in view 1. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[455.0, 242.0, 541.0, 332.0], [310.0, 618.0, 458.0, 800.0]], "image_paths": ["images/img_000074.png", "images/img_000083.png", "images/img_000081.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001171", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'desert camouflage helicopter' that are visible in view 3 but not in view 2. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[589.0, 575.0, 730.0, 688.0], [66.0, 435.0, 188.0, 513.0]], "image_paths": ["images/img_001026.png", "images/img_001017.png", "images/img_001027.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001172", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'woman in pink clothing' that are visible in view 3 but not in view 1. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[499.0, 352.0, 556.0, 477.0]], "image_paths": ["images/img_001112.png", "images/img_001124.png", "images/img_001122.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001173", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'stylized F22 fighter jet with white and red color' that are visible in view 3 but not in view 2. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[249.0, 538.0, 349.0, 615.0]], "image_paths": ["images/img_000283.png", "images/img_000281.png", "images/img_000288.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001174", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'airplane with red livery' that are visible in view 3 but not in view 1. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[623.0, 382.0, 756.0, 487.0]], "image_paths": ["images/img_000263.png", "images/img_000270.png", "images/img_000269.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001175", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'man wearing red jacket' that are visible in view 2 but not in view 3. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[409.0, 268.0, 461.0, 458.0]], "image_paths": ["images/img_000023.png", "images/img_000021.png", "images/img_000017.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001176", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'man wearing brown coat' that are visible in view 2 but not in view 1. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[480.0, 492.0, 542.0, 727.0], [701.0, 342.0, 762.0, 550.0]], "image_paths": ["images/img_000031.png", "images/img_000039.png", "images/img_000040.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001177", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'F-22 fighter jet with gray color' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[690.0, 527.0, 871.0, 620.0]], "image_paths": ["images/img_000907.png", "images/img_000902.png", "images/img_000911.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001178", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'lamp' that are visible in view 3 but not in view 2. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[585.0, 800.0, 634.0, 877.0], [688.0, 687.0, 733.0, 757.0]], "image_paths": ["images/img_000474.png", "images/img_000467.png", "images/img_000477.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001179", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'F-16 fighter jet' that are visible in view 3 but not in view 2. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[550.0, 592.0, 689.0, 720.0]], "image_paths": ["images/img_000897.png", "images/img_000885.png", "images/img_000893.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001180", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'blue shipping container' that are visible in view 2 but not in view 1. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[652.0, 423.0, 731.0, 520.0]], "image_paths": ["images/img_000483.png", "images/img_000492.png", "images/img_000491.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001181", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'snowman with red gloves and black buttons' that are visible in view 2 but not in view 3. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[456.0, 570.0, 575.0, 665.0], [574.0, 567.0, 691.0, 662.0]], "image_paths": ["images/img_000749.png", "images/img_000752.png", "images/img_000744.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001182", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'beach ball' that are visible in view 2 but not in view 1. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[271.0, 508.0, 307.0, 555.0]], "image_paths": ["images/img_000660.png", "images/img_000668.png", "images/img_000672.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001183", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'bench made of black metal' that are visible in view 2 but not in view 1. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[624.0, 582.0, 816.0, 707.0]], "image_paths": ["images/img_000632.png", "images/img_000640.png", "images/img_000638.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001184", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'man wearing gray hat' that are visible in view 3 but not in view 1. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[511.0, 197.0, 542.0, 278.0]], "image_paths": ["images/img_000755.png", "images/img_000764.png", "images/img_000765.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001185", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'green camouflage helicopter' that are visible in view 3 but not in view 2. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[686.0, 403.0, 799.0, 463.0], [720.0, 482.0, 854.0, 557.0]], "image_paths": ["images/img_001026.png", "images/img_001016.png", "images/img_001024.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001186", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'white truck with flatbed' that are visible in view 2 but not in view 1. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[636.0, 578.0, 724.0, 655.0], [689.0, 513.0, 770.0, 583.0]], "image_paths": ["images/img_000547.png", "images/img_000555.png", "images/img_000557.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001187", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'woman in white clothing' that are visible in view 1 but not in view 3. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[472.0, 488.0, 519.0, 630.0], [434.0, 455.0, 477.0, 598.0]], "image_paths": ["images/img_001056.png", "images/img_001057.png", "images/img_001049.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001188", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'white shipping container' that are visible in view 2 but not in view 3. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[246.0, 352.0, 311.0, 470.0]], "image_paths": ["images/img_000117.png", "images/img_000116.png", "images/img_000111.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001189", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'green street lamp' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[700.0, 467.0, 741.0, 592.0], [273.0, 293.0, 311.0, 413.0]], "image_paths": ["images/img_000149.png", "images/img_000145.png", "images/img_000148.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001190", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'drone with white oval body shape' that are visible in view 2 but not in view 3. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[802.0, 538.0, 843.0, 567.0], [552.0, 365.0, 584.0, 382.0], [242.0, 622.0, 301.0, 658.0], [188.0, 562.0, 226.0, 592.0], [764.0, 458.0, 807.0, 482.0]], "image_paths": ["images/img_000344.png", "images/img_000348.png", "images/img_000341.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001191", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'man wearing black coat' that are visible in view 1 but not in view 3. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[603.0, 262.0, 651.0, 462.0]], "image_paths": ["images/img_000054.png", "images/img_000052.png", "images/img_000048.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001192", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'stylized F22 fighter jet with white and red color' that are visible in view 2 but not in view 1. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[632.0, 365.0, 771.0, 467.0]], "image_paths": ["images/img_000919.png", "images/img_000927.png", "images/img_000929.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001193", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'man wearing a green hoodie and orange shoes' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[596.0, 438.0, 621.0, 535.0]], "image_paths": ["images/img_000863.png", "images/img_000857.png", "images/img_000861.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001194", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'tombstone with shoulder top' that are visible in view 3 but not in view 2. Then return their bounding boxes in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[346.0, 620.0, 422.0, 725.0]], "image_paths": ["images/img_001180.png", "images/img_001173.png", "images/img_001183.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001195", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'green shipping container' that are visible in view 1 but not in view 3. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[744.0, 600.0, 839.0, 777.0]], "image_paths": ["images/img_000090.png", "images/img_000096.png", "images/img_000087.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001196", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'white flat tombstone with cross' that are visible in view 1 but not in view 3. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[371.0, 580.0, 455.0, 752.0]], "image_paths": ["images/img_001167.png", "images/img_001163.png", "images/img_001158.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001197", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'man wearing green hood' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[414.0, 228.0, 445.0, 322.0]], "image_paths": ["images/img_000768.png", "images/img_000759.png", "images/img_000769.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001198", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'man holding box' that are visible in view 2 but not in view 1. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[338.0, 422.0, 367.0, 505.0]], "image_paths": ["images/img_000838.png", "images/img_000848.png", "images/img_000849.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001199", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'man wearing santa costume' that are visible in view 1 but not in view 2. Then return their bounding boxes in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[623.0, 567.0, 661.0, 662.0]], "image_paths": ["images/img_000802.png", "images/img_000794.png", "images/img_000796.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001200", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nLocate all 'beige sand helicopter' that are visible in view 2 but not in view 1. Then return their bounding boxes in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[390.0, 647.0, 624.0, 830.0], [550.0, 552.0, 765.0, 685.0]], "image_paths": ["images/img_000966.png", "images/img_000976.png", "images/img_000972.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001201", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'forklift' that are fully occluded by 'stack of sandbags'. For each such occluded instance of 'forklift', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[205.0, 197.0, 312.0, 357.0]], "image_paths": ["images/img_000409.png", "images/img_000401.png", "images/img_000410.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001202", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'tennis ball' that are fully occluded by 'bench made of stone'. For each such occluded instance of 'tennis ball', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[701.0, 372.0, 710.0, 383.0]], "image_paths": ["images/img_000615.png", "images/img_000624.png", "images/img_000623.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001203", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stone steps'. For each such occluded instance of 'truck', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[0.0, 405.0, 249.0, 880.0]], "image_paths": ["images/img_000475.png", "images/img_000478.png", "images/img_000469.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001204", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[318.0, 582.0, 399.0, 642.0]], "image_paths": ["images/img_000546.png", "images/img_000550.png", "images/img_000552.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001205", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'soccer ball' that are fully occluded by 'bench with armrests'. For each such occluded instance of 'soccer ball', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[774.0, 615.0, 795.0, 642.0], [786.0, 602.0, 807.0, 628.0]], "image_paths": ["images/img_000571.png", "images/img_000561.png", "images/img_000569.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001206", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'forklift' that are fully occluded by 'stack of sandbags'. For each such occluded instance of 'forklift', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[0.0, 378.0, 179.0, 528.0]], "image_paths": ["images/img_000410.png", "images/img_000401.png", "images/img_000408.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001207", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'white shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[620.0, 368.0, 651.0, 433.0]], "image_paths": ["images/img_000549.png", "images/img_000557.png", "images/img_000556.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001208", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'green shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[594.0, 343.0, 634.0, 427.0]], "image_paths": ["images/img_000483.png", "images/img_000490.png", "images/img_000491.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001209", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'forklift' that are fully occluded by 'green shipping container'. For each such occluded instance of 'forklift', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[250.0, 430.0, 302.0, 473.0]], "image_paths": ["images/img_000520.png", "images/img_000524.png", "images/img_000515.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001210", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench made of natural wood'. For each such occluded instance of 'ball', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[840.0, 450.0, 851.0, 463.0]], "image_paths": ["images/img_000604.png", "images/img_000597.png", "images/img_000607.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001211", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'white shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[491.0, 332.0, 544.0, 393.0]], "image_paths": ["images/img_000556.png", "images/img_000555.png", "images/img_000549.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001212", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'natural wooden bench with armrests'. For each such occluded instance of 'ball', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[460.0, 280.0, 468.0, 292.0]], "image_paths": ["images/img_000604.png", "images/img_000605.png", "images/img_000597.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001213", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'beach ball' that are fully occluded by 'green wooden bench with armrests'. For each such occluded instance of 'beach ball', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[787.0, 445.0, 804.0, 465.0]], "image_paths": ["images/img_000638.png", "images/img_000628.png", "images/img_000634.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001214", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'beach ball' that are fully occluded by 'bench'. For each such occluded instance of 'beach ball', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[710.0, 375.0, 724.0, 393.0]], "image_paths": ["images/img_000641.png", "images/img_000638.png", "images/img_000632.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001215", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'forklift' that are fully occluded by 'stack of sandbags'. For each such occluded instance of 'forklift', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[681.0, 192.0, 790.0, 352.0]], "image_paths": ["images/img_000401.png", "images/img_000411.png", "images/img_000409.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001216", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'green shipping container' that are fully occluded by 'white shipping container'. For each such occluded instance of 'green shipping container', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[750.0, 475.0, 836.0, 583.0]], "image_paths": ["images/img_000487.png", "images/img_000496.png", "images/img_000495.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001217", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'baseball' that are fully occluded by 'stone bench without armrests'. For each such occluded instance of 'baseball', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[636.0, 337.0, 646.0, 350.0]], "image_paths": ["images/img_000624.png", "images/img_000615.png", "images/img_000622.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001218", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[454.0, 328.0, 511.0, 393.0]], "image_paths": ["images/img_000553.png", "images/img_000546.png", "images/img_000551.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001219", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'white painted wooden bench with armrests'. For each such occluded instance of 'ball', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[329.0, 330.0, 343.0, 348.0]], "image_paths": ["images/img_000601.png", "images/img_000595.png", "images/img_000605.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001220", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'red shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[264.0, 352.0, 340.0, 397.0]], "image_paths": ["images/img_000525.png", "images/img_000518.png", "images/img_000524.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001221", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stone steps'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[679.0, 107.0, 899.0, 388.0]], "image_paths": ["images/img_000477.png", "images/img_000479.png", "images/img_000471.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001222", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench made of white painted wood'. For each such occluded instance of 'ball', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[718.0, 357.0, 733.0, 375.0]], "image_paths": ["images/img_000603.png", "images/img_000605.png", "images/img_000595.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001223", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stone steps'. For each such occluded instance of 'truck', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[0.0, 185.0, 214.0, 510.0]], "image_paths": ["images/img_000476.png", "images/img_000469.png", "images/img_000479.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001224", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'tennis ball' that are fully occluded by 'stone bench without armrests'. For each such occluded instance of 'tennis ball', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[701.0, 372.0, 710.0, 383.0]], "image_paths": ["images/img_000621.png", "images/img_000624.png", "images/img_000614.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001225", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'white shipping container' that are fully occluded by 'white shipping container'. For each such occluded instance of 'white shipping container', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[265.0, 553.0, 375.0, 622.0]], "image_paths": ["images/img_000495.png", "images/img_000493.png", "images/img_000487.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001226", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[177.0, 228.0, 315.0, 352.0]], "image_paths": ["images/img_000426.png", "images/img_000427.png", "images/img_000420.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001227", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'black stack of bricks'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[655.0, 227.0, 789.0, 345.0]], "image_paths": ["images/img_000456.png", "images/img_000458.png", "images/img_000450.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001228", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'beach ball' that are fully occluded by 'bench with armrests'. For each such occluded instance of 'beach ball', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[651.0, 312.0, 665.0, 330.0]], "image_paths": ["images/img_000587.png", "images/img_000580.png", "images/img_000589.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001229", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'worker' that are fully occluded by 'stone steps'. For each such occluded instance of 'worker', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[204.0, 567.0, 240.0, 642.0]], "image_paths": ["images/img_000469.png", "images/img_000476.png", "images/img_000475.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001230", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'worker' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'worker', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[468.0, 268.0, 487.0, 333.0]], "image_paths": ["images/img_000424.png", "images/img_000425.png", "images/img_000416.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001231", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'soccer ball' that are fully occluded by 'bench'. For each such occluded instance of 'soccer ball', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[444.0, 268.0, 458.0, 287.0], [135.0, 448.0, 154.0, 472.0]], "image_paths": ["images/img_000632.png", "images/img_000638.png", "images/img_000639.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001232", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'worker' that are fully occluded by 'stack of sandbags'. For each such occluded instance of 'worker', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[708.0, 328.0, 730.0, 397.0]], "image_paths": ["images/img_000441.png", "images/img_000439.png", "images/img_000431.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001233", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[292.0, 333.0, 376.0, 435.0]], "image_paths": ["images/img_000533.png", "images/img_000541.png", "images/img_000540.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001234", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[480.0, 260.0, 546.0, 312.0]], "image_paths": ["images/img_000493.png", "images/img_000484.png", "images/img_000492.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001235", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'concrete mixer' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'concrete mixer', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[838.0, 417.0, 892.0, 482.0], [802.0, 395.0, 849.0, 463.0]], "image_paths": ["images/img_000440.png", "images/img_000432.png", "images/img_000441.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001236", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'natural wooden bench with armrests'. For each such occluded instance of 'ball', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[109.0, 692.0, 130.0, 718.0]], "image_paths": ["images/img_000570.png", "images/img_000564.png", "images/img_000572.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001237", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'soccer ball' that are fully occluded by 'bench with armrests'. For each such occluded instance of 'soccer ball', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[453.0, 303.0, 466.0, 322.0], [438.0, 303.0, 451.0, 322.0]], "image_paths": ["images/img_000561.png", "images/img_000568.png", "images/img_000571.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001238", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[347.0, 367.0, 381.0, 435.0]], "image_paths": ["images/img_000551.png", "images/img_000552.png", "images/img_000546.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001239", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'baseball' that are fully occluded by 'green wooden bench with armrests'. For each such occluded instance of 'baseball', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[549.0, 308.0, 559.0, 322.0]], "image_paths": ["images/img_000559.png", "images/img_000569.png", "images/img_000568.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001240", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck with white and black cab' that are fully occluded by 'red shipping container'. For each such occluded instance of 'truck with white and black cab', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[292.0, 333.0, 376.0, 435.0]], "image_paths": ["images/img_000534.png", "images/img_000541.png", "images/img_000543.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001241", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[576.0, 308.0, 657.0, 405.0]], "image_paths": ["images/img_000542.png", "images/img_000543.png", "images/img_000533.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001242", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'worker' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'worker', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[468.0, 268.0, 487.0, 333.0]], "image_paths": ["images/img_000416.png", "images/img_000424.png", "images/img_000423.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001243", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'worker' that are fully occluded by 'stack of sandbags'. For each such occluded instance of 'worker', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[211.0, 400.0, 237.0, 465.0]], "image_paths": ["images/img_000438.png", "images/img_000441.png", "images/img_000431.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001244", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'soccer ball' that are fully occluded by 'green wooden bench with armrests'. For each such occluded instance of 'soccer ball', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[453.0, 303.0, 466.0, 322.0], [438.0, 303.0, 451.0, 322.0]], "image_paths": ["images/img_000570.png", "images/img_000560.png", "images/img_000568.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001245", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'baseball' that are fully occluded by 'bench made of stone'. For each such occluded instance of 'baseball', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[429.0, 318.0, 439.0, 332.0]], "image_paths": ["images/img_000582.png", "images/img_000589.png", "images/img_000590.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001246", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'trolley' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'trolley', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[180.0, 432.0, 249.0, 475.0]], "image_paths": ["images/img_000413.png", "images/img_000412.png", "images/img_000406.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001247", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'soccer ball' that are fully occluded by 'stone bench without armrests'. For each such occluded instance of 'soccer ball', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[646.0, 337.0, 661.0, 355.0], [672.0, 350.0, 688.0, 370.0]], "image_paths": ["images/img_000623.png", "images/img_000624.png", "images/img_000616.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001248", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'tennis ball' that are fully occluded by 'green wooden bench with armrests'. For each such occluded instance of 'tennis ball', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[814.0, 593.0, 826.0, 610.0], [824.0, 580.0, 836.0, 595.0]], "image_paths": ["images/img_000637.png", "images/img_000639.png", "images/img_000629.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001249", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench made of white painted wood'. For each such occluded instance of 'ball', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[792.0, 513.0, 810.0, 535.0]], "image_paths": ["images/img_000602.png", "images/img_000604.png", "images/img_000595.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001250", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'worker wearing face mask and red shirt' that are fully occluded by 'stone steps'. For each such occluded instance of 'worker wearing face mask and red shirt', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[117.0, 392.0, 154.0, 468.0]], "image_paths": ["images/img_000402.png", "images/img_000412.png", "images/img_000409.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001251", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stack of sandbags'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[843.0, 418.0, 1000.0, 543.0]], "image_paths": ["images/img_000433.png", "images/img_000442.png", "images/img_000440.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001252", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench made of natural wood'. For each such occluded instance of 'ball', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[226.0, 348.0, 236.0, 360.0]], "image_paths": ["images/img_000607.png", "images/img_000597.png", "images/img_000604.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001253", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'concrete mixer' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'concrete mixer', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[736.0, 362.0, 780.0, 430.0]], "image_paths": ["images/img_000426.png", "images/img_000425.png", "images/img_000417.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001254", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'white shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[165.0, 482.0, 255.0, 552.0]], "image_paths": ["images/img_000523.png", "images/img_000524.png", "images/img_000517.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001255", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'black stack of bricks'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[0.0, 430.0, 164.0, 555.0]], "image_paths": ["images/img_000456.png", "images/img_000455.png", "images/img_000449.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001256", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'worker' that are fully occluded by 'black stack of bricks'. For each such occluded instance of 'worker', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[537.0, 275.0, 561.0, 332.0]], "image_paths": ["images/img_000455.png", "images/img_000449.png", "images/img_000457.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001257", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stack of sandbags'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[762.0, 667.0, 1000.0, 937.0]], "image_paths": ["images/img_000441.png", "images/img_000443.png", "images/img_000433.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001258", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'green shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[652.0, 423.0, 731.0, 520.0]], "image_paths": ["images/img_000492.png", "images/img_000483.png", "images/img_000491.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001259", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stone steps'. For each such occluded instance of 'truck', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[0.0, 405.0, 249.0, 880.0]], "image_paths": ["images/img_000475.png", "images/img_000476.png", "images/img_000469.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001260", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'worker' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'worker', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[657.0, 305.0, 682.0, 370.0]], "image_paths": ["images/img_000416.png", "images/img_000424.png", "images/img_000425.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001261", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench made of natural wood'. For each such occluded instance of 'ball', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[460.0, 280.0, 468.0, 292.0]], "image_paths": ["images/img_000604.png", "images/img_000598.png", "images/img_000605.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001262", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'worker' that are fully occluded by 'stone steps'. For each such occluded instance of 'worker', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[219.0, 387.0, 251.0, 457.0]], "image_paths": ["images/img_000479.png", "images/img_000470.png", "images/img_000476.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001263", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'worker' that are fully occluded by 'black stack of bricks'. For each such occluded instance of 'worker', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[537.0, 275.0, 561.0, 332.0]], "image_paths": ["images/img_000449.png", "images/img_000455.png", "images/img_000457.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001264", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stone steps'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[471.0, 73.0, 593.0, 315.0]], "image_paths": ["images/img_000469.png", "images/img_000478.png", "images/img_000477.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001265", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck with blue and white cab' that are fully occluded by 'red shipping container'. For each such occluded instance of 'truck with blue and white cab', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[626.0, 302.0, 674.0, 383.0]], "image_paths": ["images/img_000534.png", "images/img_000542.png", "images/img_000543.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001266", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'trolley' that are fully occluded by 'stone steps'. For each such occluded instance of 'trolley', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[120.0, 422.0, 186.0, 470.0]], "image_paths": ["images/img_000417.png", "images/img_000423.png", "images/img_000426.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001267", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[271.0, 460.0, 336.0, 542.0]], "image_paths": ["images/img_000551.png", "images/img_000546.png", "images/img_000550.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001268", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[347.0, 367.0, 381.0, 435.0]], "image_paths": ["images/img_000552.png", "images/img_000546.png", "images/img_000554.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001269", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'beach ball' that are fully occluded by 'bench made of blue painted metal'. For each such occluded instance of 'beach ball', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[221.0, 392.0, 236.0, 412.0]], "image_paths": ["images/img_000577.png", "images/img_000587.png", "images/img_000583.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001270", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'baseball' that are fully occluded by 'bench made of blue painted metal'. For each such occluded instance of 'baseball', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[764.0, 408.0, 775.0, 423.0]], "image_paths": ["images/img_000577.png", "images/img_000587.png", "images/img_000586.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001271", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench made of white painted wood'. For each such occluded instance of 'ball', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[792.0, 513.0, 810.0, 535.0]], "image_paths": ["images/img_000601.png", "images/img_000604.png", "images/img_000595.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001272", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'baseball' that are fully occluded by 'blue painted metal bench without armrests'. For each such occluded instance of 'baseball', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[806.0, 583.0, 821.0, 602.0]], "image_paths": ["images/img_000586.png", "images/img_000577.png", "images/img_000587.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001273", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'black stack of bricks'. For each such occluded instance of 'truck', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[166.0, 250.0, 301.0, 373.0]], "image_paths": ["images/img_000456.png", "images/img_000450.png", "images/img_000458.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001274", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'trolley' that are fully occluded by 'stone steps'. For each such occluded instance of 'trolley', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[110.0, 653.0, 164.0, 715.0]], "image_paths": ["images/img_000417.png", "images/img_000424.png", "images/img_000422.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001275", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stone steps'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[0.0, 185.0, 214.0, 510.0]], "image_paths": ["images/img_000469.png", "images/img_000476.png", "images/img_000479.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001276", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'soccer ball' that are fully occluded by 'green wooden bench with armrests'. For each such occluded instance of 'soccer ball', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[584.0, 292.0, 599.0, 312.0], [474.0, 293.0, 489.0, 313.0]], "image_paths": ["images/img_000635.png", "images/img_000628.png", "images/img_000636.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001277", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'concrete mixer' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'concrete mixer', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[838.0, 417.0, 892.0, 482.0], [802.0, 395.0, 849.0, 463.0]], "image_paths": ["images/img_000440.png", "images/img_000431.png", "images/img_000441.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001278", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'white shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[625.0, 635.0, 711.0, 702.0]], "image_paths": ["images/img_000515.png", "images/img_000522.png", "images/img_000524.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001279", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'forklift' that are fully occluded by 'green shipping container'. For each such occluded instance of 'forklift', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[366.0, 340.0, 403.0, 387.0]], "image_paths": ["images/img_000515.png", "images/img_000521.png", "images/img_000520.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001280", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'red shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[438.0, 297.0, 499.0, 345.0]], "image_paths": ["images/img_000525.png", "images/img_000527.png", "images/img_000518.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001281", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[807.0, 457.0, 910.0, 535.0]], "image_paths": ["images/img_000494.png", "images/img_000491.png", "images/img_000484.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001282", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'beach ball' that are fully occluded by 'bench made of stone'. For each such occluded instance of 'beach ball', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[198.0, 490.0, 214.0, 510.0]], "image_paths": ["images/img_000621.png", "images/img_000622.png", "images/img_000615.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001283", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench'. For each such occluded instance of 'ball', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[835.0, 438.0, 853.0, 460.0]], "image_paths": ["images/img_000631.png", "images/img_000641.png", "images/img_000637.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001284", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[292.0, 333.0, 376.0, 435.0]], "image_paths": ["images/img_000540.png", "images/img_000533.png", "images/img_000541.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001285", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'black stack of bricks'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[441.0, 190.0, 514.0, 293.0]], "image_paths": ["images/img_000458.png", "images/img_000457.png", "images/img_000450.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001286", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'beach ball' that are fully occluded by 'bench'. For each such occluded instance of 'beach ball', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[220.0, 437.0, 236.0, 457.0]], "image_paths": ["images/img_000638.png", "images/img_000640.png", "images/img_000632.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001287", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'white shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[511.0, 305.0, 568.0, 368.0]], "image_paths": ["images/img_000520.png", "images/img_000521.png", "images/img_000515.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001288", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'worker' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'worker', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[468.0, 268.0, 487.0, 333.0]], "image_paths": ["images/img_000418.png", "images/img_000424.png", "images/img_000425.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001289", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stone steps'. For each such occluded instance of 'truck', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[155.0, 88.0, 370.0, 360.0]], "image_paths": ["images/img_000475.png", "images/img_000469.png", "images/img_000477.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001290", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[460.0, 168.0, 545.0, 298.0]], "image_paths": ["images/img_000407.png", "images/img_000414.png", "images/img_000415.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001291", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench made of natural wood'. For each such occluded instance of 'ball', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[460.0, 280.0, 468.0, 292.0]], "image_paths": ["images/img_000598.png", "images/img_000605.png", "images/img_000607.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001292", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'trolley' that are fully occluded by 'stone steps'. For each such occluded instance of 'trolley', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[120.0, 422.0, 186.0, 470.0]], "image_paths": ["images/img_000423.png", "images/img_000426.png", "images/img_000418.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001293", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'forklift' that are fully occluded by 'stack of sandbags'. For each such occluded instance of 'forklift', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[819.0, 362.0, 1000.0, 512.0]], "image_paths": ["images/img_000412.png", "images/img_000402.png", "images/img_000411.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001294", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[119.0, 437.0, 194.0, 513.0]], "image_paths": ["images/img_000484.png", "images/img_000494.png", "images/img_000490.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001295", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'worker' that are fully occluded by 'stack of sandbags'. For each such occluded instance of 'worker', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[354.0, 290.0, 375.0, 350.0]], "image_paths": ["images/img_000439.png", "images/img_000441.png", "images/img_000433.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001296", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'soccer ball' that are fully occluded by 'green wooden bench with armrests'. For each such occluded instance of 'soccer ball', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[830.0, 565.0, 851.0, 590.0], [792.0, 457.0, 811.0, 480.0]], "image_paths": ["images/img_000628.png", "images/img_000638.png", "images/img_000635.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001297", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'red shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[626.0, 322.0, 659.0, 375.0]], "image_paths": ["images/img_000518.png", "images/img_000526.png", "images/img_000525.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001298", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[206.0, 450.0, 321.0, 542.0]], "image_paths": ["images/img_000533.png", "images/img_000541.png", "images/img_000540.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001299", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[675.0, 312.0, 757.0, 360.0]], "image_paths": ["images/img_000484.png", "images/img_000493.png", "images/img_000494.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001300", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'beach ball' that are fully occluded by 'bench made of blue painted metal'. For each such occluded instance of 'beach ball', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[221.0, 392.0, 236.0, 412.0]], "image_paths": ["images/img_000583.png", "images/img_000587.png", "images/img_000577.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001301", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[206.0, 450.0, 321.0, 542.0]], "image_paths": ["images/img_000539.png", "images/img_000540.png", "images/img_000533.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001302", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'tennis ball' that are fully occluded by 'bench made of stone'. For each such occluded instance of 'tennis ball', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[529.0, 310.0, 537.0, 322.0]], "image_paths": ["images/img_000615.png", "images/img_000623.png", "images/img_000621.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001303", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'basketball' that are fully occluded by 'bench made of blue painted metal'. For each such occluded instance of 'basketball', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[733.0, 352.0, 749.0, 373.0], [736.0, 362.0, 754.0, 383.0]], "image_paths": ["images/img_000570.png", "images/img_000571.png", "images/img_000561.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001304", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'white shipping container' that are fully occluded by 'white shipping container'. For each such occluded instance of 'white shipping container', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[369.0, 338.0, 407.0, 423.0]], "image_paths": ["images/img_000487.png", "images/img_000495.png", "images/img_000493.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001305", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[292.0, 333.0, 376.0, 435.0]], "image_paths": ["images/img_000541.png", "images/img_000533.png", "images/img_000539.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001306", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'soccer ball' that are fully occluded by 'green wooden bench with armrests'. For each such occluded instance of 'soccer ball', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[760.0, 433.0, 777.0, 455.0], [756.0, 420.0, 774.0, 442.0]], "image_paths": ["images/img_000560.png", "images/img_000569.png", "images/img_000570.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001307", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'natural wooden bench with armrests'. For each such occluded instance of 'ball', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[840.0, 450.0, 851.0, 463.0]], "image_paths": ["images/img_000605.png", "images/img_000607.png", "images/img_000598.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001308", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'white shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[556.0, 300.0, 609.0, 345.0]], "image_paths": ["images/img_000495.png", "images/img_000488.png", "images/img_000496.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001309", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'worker' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'worker', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[468.0, 268.0, 487.0, 333.0]], "image_paths": ["images/img_000424.png", "images/img_000417.png", "images/img_000423.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001310", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'truck', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[159.0, 210.0, 336.0, 358.0]], "image_paths": ["images/img_000406.png", "images/img_000414.png", "images/img_000413.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001311", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench with armrests'. For each such occluded instance of 'ball', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[250.0, 382.0, 268.0, 403.0]], "image_paths": ["images/img_000612.png", "images/img_000619.png", "images/img_000618.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001312", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'basketball' that are fully occluded by 'bench with armrests'. For each such occluded instance of 'basketball', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[716.0, 373.0, 734.0, 395.0], [662.0, 345.0, 679.0, 367.0]], "image_paths": ["images/img_000627.png", "images/img_000635.png", "images/img_000637.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001313", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'white shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[491.0, 332.0, 544.0, 393.0]], "image_paths": ["images/img_000549.png", "images/img_000555.png", "images/img_000557.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001314", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[271.0, 460.0, 336.0, 542.0]], "image_paths": ["images/img_000552.png", "images/img_000551.png", "images/img_000546.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001315", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'white shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[165.0, 482.0, 255.0, 552.0]], "image_paths": ["images/img_000517.png", "images/img_000523.png", "images/img_000524.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001316", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'beach ball' that are fully occluded by 'stone bench without armrests'. For each such occluded instance of 'beach ball', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[660.0, 345.0, 674.0, 362.0]], "image_paths": ["images/img_000624.png", "images/img_000614.png", "images/img_000622.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001317", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'white shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[776.0, 437.0, 891.0, 518.0]], "image_paths": ["images/img_000537.png", "images/img_000538.png", "images/img_000529.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001318", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'worker' that are fully occluded by 'black stack of bricks'. For each such occluded instance of 'worker', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[215.0, 397.0, 242.0, 470.0]], "image_paths": ["images/img_000457.png", "images/img_000455.png", "images/img_000449.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001319", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench made of natural wood'. For each such occluded instance of 'ball', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[226.0, 348.0, 236.0, 360.0]], "image_paths": ["images/img_000604.png", "images/img_000605.png", "images/img_000598.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001320", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench with armrests'. For each such occluded instance of 'ball', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[795.0, 608.0, 818.0, 637.0]], "image_paths": ["images/img_000622.png", "images/img_000612.png", "images/img_000619.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001321", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'trolley' that are fully occluded by 'stone steps'. For each such occluded instance of 'trolley', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[859.0, 483.0, 912.0, 552.0]], "image_paths": ["images/img_000424.png", "images/img_000427.png", "images/img_000418.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001322", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck with white and black cab' that are fully occluded by 'green shipping container'. For each such occluded instance of 'truck with white and black cab', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[136.0, 645.0, 282.0, 757.0]], "image_paths": ["images/img_000534.png", "images/img_000543.png", "images/img_000540.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001323", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'beach ball' that are fully occluded by 'bench made of stone'. For each such occluded instance of 'beach ball', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[326.0, 352.0, 340.0, 370.0]], "image_paths": ["images/img_000623.png", "images/img_000615.png", "images/img_000622.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001324", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'basketball' that are fully occluded by 'blue painted metal bench without armrests'. For each such occluded instance of 'basketball', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[850.0, 528.0, 871.0, 555.0], [840.0, 540.0, 861.0, 567.0]], "image_paths": ["images/img_000570.png", "images/img_000571.png", "images/img_000561.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001325", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'black stack of bricks'. For each such occluded instance of 'truck', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[441.0, 190.0, 514.0, 293.0]], "image_paths": ["images/img_000450.png", "images/img_000459.png", "images/img_000457.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001326", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'green shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[652.0, 423.0, 731.0, 520.0]], "image_paths": ["images/img_000483.png", "images/img_000492.png", "images/img_000489.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001327", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench made of natural wood'. For each such occluded instance of 'ball', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[460.0, 280.0, 468.0, 292.0]], "image_paths": ["images/img_000605.png", "images/img_000597.png", "images/img_000603.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001328", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'natural wooden bench with armrests'. For each such occluded instance of 'ball', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[105.0, 517.0, 117.0, 532.0]], "image_paths": ["images/img_000607.png", "images/img_000603.png", "images/img_000597.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001329", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'white truck with flatbed' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'white truck with flatbed', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[111.0, 662.0, 239.0, 738.0]], "image_paths": ["images/img_000483.png", "images/img_000489.png", "images/img_000491.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001330", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'forklift' that are fully occluded by 'green shipping container'. For each such occluded instance of 'forklift', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[656.0, 373.0, 693.0, 423.0]], "image_paths": ["images/img_000520.png", "images/img_000515.png", "images/img_000523.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001331", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck with blue and white cab' that are fully occluded by 'red shipping container'. For each such occluded instance of 'truck with blue and white cab', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[734.0, 403.0, 818.0, 505.0]], "image_paths": ["images/img_000540.png", "images/img_000543.png", "images/img_000534.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001332", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'soccer ball' that are fully occluded by 'bench with armrests'. For each such occluded instance of 'soccer ball', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[760.0, 433.0, 777.0, 455.0], [756.0, 420.0, 774.0, 442.0]], "image_paths": ["images/img_000570.png", "images/img_000569.png", "images/img_000561.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001333", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'soccer ball' that are fully occluded by 'stone bench without armrests'. For each such occluded instance of 'soccer ball', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[443.0, 312.0, 458.0, 332.0]], "image_paths": ["images/img_000589.png", "images/img_000582.png", "images/img_000590.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001334", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stone steps'. For each such occluded instance of 'truck', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[679.0, 107.0, 899.0, 388.0]], "image_paths": ["images/img_000476.png", "images/img_000469.png", "images/img_000479.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001335", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'worker' that are fully occluded by 'stack of sandbags'. For each such occluded instance of 'worker', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[517.0, 258.0, 531.0, 323.0]], "image_paths": ["images/img_000410.png", "images/img_000401.png", "images/img_000408.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001336", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'soccer ball' that are fully occluded by 'bench with armrests'. For each such occluded instance of 'soccer ball', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[774.0, 615.0, 795.0, 642.0], [786.0, 602.0, 807.0, 628.0]], "image_paths": ["images/img_000571.png", "images/img_000568.png", "images/img_000561.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001337", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[576.0, 308.0, 657.0, 405.0]], "image_paths": ["images/img_000533.png", "images/img_000543.png", "images/img_000542.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001338", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'soccer ball' that are fully occluded by 'bench made of stone'. For each such occluded instance of 'soccer ball', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[274.0, 358.0, 290.0, 380.0]], "image_paths": ["images/img_000590.png", "images/img_000582.png", "images/img_000588.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001339", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stone steps'. For each such occluded instance of 'truck', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[0.0, 185.0, 214.0, 510.0]], "image_paths": ["images/img_000469.png", "images/img_000479.png", "images/img_000476.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001340", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'white shipping container' that are fully occluded by 'white shipping container'. For each such occluded instance of 'white shipping container', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[481.0, 317.0, 556.0, 397.0]], "image_paths": ["images/img_000496.png", "images/img_000495.png", "images/img_000487.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001341", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'tennis ball' that are fully occluded by 'bench with armrests'. For each such occluded instance of 'tennis ball', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[519.0, 310.0, 527.0, 322.0]], "image_paths": ["images/img_000568.png", "images/img_000571.png", "images/img_000561.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001342", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'worker' that are fully occluded by 'stack of sandbags'. For each such occluded instance of 'worker', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[708.0, 328.0, 730.0, 397.0]], "image_paths": ["images/img_000439.png", "images/img_000441.png", "images/img_000433.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001343", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'white shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[666.0, 462.0, 728.0, 538.0]], "image_paths": ["images/img_000549.png", "images/img_000557.png", "images/img_000555.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001344", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stone steps'. For each such occluded instance of 'truck', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[155.0, 88.0, 370.0, 360.0]], "image_paths": ["images/img_000471.png", "images/img_000479.png", "images/img_000477.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001345", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'worker' that are fully occluded by 'stack of sandbags'. For each such occluded instance of 'worker', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[708.0, 328.0, 730.0, 397.0]], "image_paths": ["images/img_000441.png", "images/img_000438.png", "images/img_000431.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001346", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'trolley' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'trolley', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[321.0, 308.0, 367.0, 358.0]], "image_paths": ["images/img_000415.png", "images/img_000406.png", "images/img_000413.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001347", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stone steps'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[471.0, 73.0, 593.0, 315.0]], "image_paths": ["images/img_000469.png", "images/img_000478.png", "images/img_000476.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001348", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'tennis ball' that are fully occluded by 'bench made of stone'. For each such occluded instance of 'tennis ball', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[215.0, 457.0, 225.0, 470.0]], "image_paths": ["images/img_000620.png", "images/img_000614.png", "images/img_000621.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001349", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[581.0, 363.0, 644.0, 408.0]], "image_paths": ["images/img_000551.png", "images/img_000554.png", "images/img_000546.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001350", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'basketball' that are fully occluded by 'bench with armrests'. For each such occluded instance of 'basketball', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[792.0, 418.0, 812.0, 443.0]], "image_paths": ["images/img_000576.png", "images/img_000583.png", "images/img_000586.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001351", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[581.0, 363.0, 644.0, 408.0]], "image_paths": ["images/img_000546.png", "images/img_000551.png", "images/img_000554.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001352", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'white shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[263.0, 647.0, 316.0, 752.0]], "image_paths": ["images/img_000523.png", "images/img_000517.png", "images/img_000522.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001353", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'red shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[626.0, 322.0, 659.0, 375.0]], "image_paths": ["images/img_000526.png", "images/img_000518.png", "images/img_000527.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001354", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'baseball' that are fully occluded by 'bench made of stone'. For each such occluded instance of 'baseball', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[263.0, 373.0, 274.0, 388.0]], "image_paths": ["images/img_000582.png", "images/img_000588.png", "images/img_000590.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001355", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'forklift' that are fully occluded by 'stack of sandbags'. For each such occluded instance of 'forklift', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[819.0, 362.0, 1000.0, 512.0]], "image_paths": ["images/img_000403.png", "images/img_000413.png", "images/img_000412.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001356", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench'. For each such occluded instance of 'ball', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[352.0, 323.0, 366.0, 342.0]], "image_paths": ["images/img_000640.png", "images/img_000631.png", "images/img_000639.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001357", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'soccer ball' that are fully occluded by 'green wooden bench with armrests'. For each such occluded instance of 'soccer ball', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[627.0, 327.0, 642.0, 347.0], [615.0, 320.0, 630.0, 340.0]], "image_paths": ["images/img_000560.png", "images/img_000569.png", "images/img_000566.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001358", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'concrete mixer' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'concrete mixer', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[247.0, 298.0, 269.0, 360.0], [237.0, 325.0, 266.0, 387.0]], "image_paths": ["images/img_000438.png", "images/img_000432.png", "images/img_000439.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001359", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'black stack of bricks'. For each such occluded instance of 'truck', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[441.0, 190.0, 514.0, 293.0]], "image_paths": ["images/img_000457.png", "images/img_000459.png", "images/img_000450.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001360", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'worker' that are fully occluded by 'stack of sandbags'. For each such occluded instance of 'worker', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[211.0, 400.0, 237.0, 465.0]], "image_paths": ["images/img_000438.png", "images/img_000439.png", "images/img_000432.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001361", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'basketball' that are fully occluded by 'blue painted metal bench without armrests'. For each such occluded instance of 'basketball', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[224.0, 377.0, 241.0, 400.0], [179.0, 510.0, 200.0, 538.0], [194.0, 465.0, 214.0, 490.0], [168.0, 542.0, 190.0, 570.0]], "image_paths": ["images/img_000587.png", "images/img_000583.png", "images/img_000577.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001362", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[292.0, 333.0, 376.0, 435.0]], "image_paths": ["images/img_000543.png", "images/img_000541.png", "images/img_000533.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001363", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[347.0, 367.0, 381.0, 435.0]], "image_paths": ["images/img_000553.png", "images/img_000546.png", "images/img_000552.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001364", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'soccer ball' that are fully occluded by 'stone bench without armrests'. For each such occluded instance of 'soccer ball', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[236.0, 648.0, 258.0, 675.0]], "image_paths": ["images/img_000622.png", "images/img_000620.png", "images/img_000614.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001365", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'trolley' that are fully occluded by 'stone steps'. For each such occluded instance of 'trolley', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[733.0, 313.0, 762.0, 370.0]], "image_paths": ["images/img_000418.png", "images/img_000427.png", "images/img_000426.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001366", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'basketball' that are fully occluded by 'green wooden bench with armrests'. For each such occluded instance of 'basketball', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[259.0, 375.0, 276.0, 397.0], [316.0, 345.0, 333.0, 367.0]], "image_paths": ["images/img_000635.png", "images/img_000629.png", "images/img_000639.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001367", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'forklift' that are fully occluded by 'stack of sandbags'. For each such occluded instance of 'forklift', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[205.0, 197.0, 312.0, 357.0]], "image_paths": ["images/img_000409.png", "images/img_000412.png", "images/img_000402.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001368", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[441.0, 190.0, 514.0, 293.0]], "image_paths": ["images/img_000458.png", "images/img_000457.png", "images/img_000448.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001369", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'white shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[264.0, 352.0, 340.0, 397.0]], "image_paths": ["images/img_000517.png", "images/img_000524.png", "images/img_000522.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001370", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[284.0, 302.0, 324.0, 360.0]], "image_paths": ["images/img_000484.png", "images/img_000491.png", "images/img_000490.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001371", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stack of sandbags'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[843.0, 418.0, 1000.0, 543.0]], "image_paths": ["images/img_000440.png", "images/img_000442.png", "images/img_000433.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001372", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[119.0, 437.0, 194.0, 513.0]], "image_paths": ["images/img_000484.png", "images/img_000492.png", "images/img_000490.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001373", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'beach ball' that are fully occluded by 'green wooden bench with armrests'. For each such occluded instance of 'beach ball', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[459.0, 295.0, 471.0, 312.0]], "image_paths": ["images/img_000636.png", "images/img_000628.png", "images/img_000634.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001374", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'forklift' that are fully occluded by 'green shipping container'. For each such occluded instance of 'forklift', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[527.0, 320.0, 542.0, 365.0]], "image_paths": ["images/img_000523.png", "images/img_000515.png", "images/img_000522.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001375", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[581.0, 363.0, 644.0, 408.0]], "image_paths": ["images/img_000554.png", "images/img_000546.png", "images/img_000553.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001376", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'ball' that are fully occluded by 'bench'. For each such occluded instance of 'ball', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[664.0, 305.0, 677.0, 323.0]], "image_paths": ["images/img_000631.png", "images/img_000640.png", "images/img_000641.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001377", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'trolley' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'trolley', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[321.0, 308.0, 367.0, 358.0]], "image_paths": ["images/img_000413.png", "images/img_000415.png", "images/img_000405.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001378", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[347.0, 367.0, 381.0, 435.0]], "image_paths": ["images/img_000546.png", "images/img_000550.png", "images/img_000552.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001379", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'worker' that are fully occluded by 'stack of sandbags'. For each such occluded instance of 'worker', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[211.0, 400.0, 237.0, 465.0]], "image_paths": ["images/img_000438.png", "images/img_000431.png", "images/img_000441.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001380", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'truck with white and black cab' that are fully occluded by 'red shipping container'. For each such occluded instance of 'truck with white and black cab', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[576.0, 308.0, 657.0, 405.0]], "image_paths": ["images/img_000540.png", "images/img_000534.png", "images/img_000543.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001381", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'basketball' that are fully occluded by 'bench with armrests'. For each such occluded instance of 'basketball', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[250.0, 382.0, 268.0, 403.0]], "image_paths": ["images/img_000612.png", "images/img_000618.png", "images/img_000621.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001382", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'worker' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'worker', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[215.0, 397.0, 242.0, 470.0]], "image_paths": ["images/img_000455.png", "images/img_000456.png", "images/img_000448.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001383", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'forklift' that are fully occluded by 'green shipping container'. For each such occluded instance of 'forklift', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[250.0, 430.0, 302.0, 473.0]], "image_paths": ["images/img_000515.png", "images/img_000524.png", "images/img_000520.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001384", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'black stack of bricks'. For each such occluded instance of 'truck', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[441.0, 190.0, 514.0, 293.0]], "image_paths": ["images/img_000457.png", "images/img_000456.png", "images/img_000450.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001385", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'white shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[355.0, 333.0, 414.0, 378.0]], "image_paths": ["images/img_000524.png", "images/img_000520.png", "images/img_000515.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001386", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'truck with white and black cab' that are fully occluded by 'green shipping container'. For each such occluded instance of 'truck with white and black cab', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[266.0, 287.0, 312.0, 383.0]], "image_paths": ["images/img_000542.png", "images/img_000534.png", "images/img_000540.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001387", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'green shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[652.0, 423.0, 731.0, 520.0]], "image_paths": ["images/img_000490.png", "images/img_000492.png", "images/img_000483.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001388", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'basketball' that are fully occluded by 'bench with armrests'. For each such occluded instance of 'basketball', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[426.0, 312.0, 441.0, 332.0], [411.0, 312.0, 426.0, 332.0]], "image_paths": ["images/img_000621.png", "images/img_000619.png", "images/img_000611.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001389", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'black stack of bricks'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[166.0, 250.0, 301.0, 373.0]], "image_paths": ["images/img_000459.png", "images/img_000456.png", "images/img_000449.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001390", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'white shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[718.0, 478.0, 785.0, 565.0]], "image_paths": ["images/img_000515.png", "images/img_000521.png", "images/img_000523.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001391", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'worker' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'worker', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[296.0, 318.0, 315.0, 383.0]], "image_paths": ["images/img_000411.png", "images/img_000413.png", "images/img_000405.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001392", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'shipping container'. For each such occluded instance of 'truck', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[450.0, 287.0, 486.0, 383.0]], "image_paths": ["images/img_000542.png", "images/img_000543.png", "images/img_000533.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001393", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'tennis ball' that are fully occluded by 'green wooden bench with armrests'. For each such occluded instance of 'tennis ball', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[782.0, 495.0, 794.0, 510.0], [802.0, 558.0, 815.0, 573.0]], "image_paths": ["images/img_000560.png", "images/img_000566.png", "images/img_000570.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001394", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'shipping container' that are fully occluded by 'blue shipping container'. For each such occluded instance of 'shipping container', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[347.0, 367.0, 381.0, 435.0]], "image_paths": ["images/img_000552.png", "images/img_000550.png", "images/img_000546.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001395", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 3, there may be zero, one, or multiple instances of 'worker' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'worker', return its corresponding bounding box in view 1. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[657.0, 305.0, 682.0, 370.0]], "image_paths": ["images/img_000425.png", "images/img_000426.png", "images/img_000418.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001396", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 2, there may be zero, one, or multiple instances of 'white shipping container' that are fully occluded by 'white shipping container'. For each such occluded instance of 'white shipping container', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[261.0, 420.0, 344.0, 518.0]], "image_paths": ["images/img_000496.png", "images/img_000487.png", "images/img_000494.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001397", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck with white and black cab' that are fully occluded by 'red shipping container'. For each such occluded instance of 'truck with white and black cab', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[292.0, 333.0, 376.0, 435.0]], "image_paths": ["images/img_000534.png", "images/img_000541.png", "images/img_000540.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001398", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'soccer ball' that are fully occluded by 'bench made of stone'. For each such occluded instance of 'soccer ball', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[274.0, 358.0, 290.0, 380.0]], "image_paths": ["images/img_000582.png", "images/img_000589.png", "images/img_000588.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001399", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'basketball' that are fully occluded by 'blue painted metal bench without armrests'. For each such occluded instance of 'basketball', return its corresponding bounding box in view 3. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[815.0, 612.0, 839.0, 643.0], [761.0, 448.0, 781.0, 473.0], [775.0, 490.0, 796.0, 517.0], [752.0, 422.0, 771.0, 447.0]], "image_paths": ["images/img_000577.png", "images/img_000583.png", "images/img_000587.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001400", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different camera positions and viewing angles.\nIn view 1, there may be zero, one, or multiple instances of 'truck' that are fully occluded by 'stack of bricks'. For each such occluded instance of 'truck', return its corresponding bounding box in view 2. \n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[0.0, 658.0, 211.0, 935.0]], "image_paths": ["images/img_000419.png", "images/img_000425.png", "images/img_000428.png"], "supporting": {"draw_boxes": null}} +{"q_idx": "detect_001401", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('woman in gray clothing'). For every instance, it has 1 woman in gray clothing to its left, and 1 food stall with red and white striped roof to its left-behind in view 1.\nIn view 2, locate all target instances 'food stall with red and white striped roof' that lie in the left-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[613.0, 258.0, 746.0, 433.0]], "image_paths": ["images/img_001079.png", "images/img_001088.png"], "supporting": {"draw_boxes": {"images/img_001079.png": []}}} +{"q_idx": "detect_001402", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('lamp'). For every instance, it has 1 forklift to its right-behind in view 3.\nIn view 2, locate all target instances 'forklift' that lie in the right-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[39.0, 637.0, 271.0, 893.0]], "image_paths": ["images/img_000438.png", "images/img_000441.png", "images/img_000435.png"], "supporting": {"draw_boxes": {"images/img_000435.png": []}}} +{"q_idx": "detect_001403", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('white upright tombstone with cross'). For every instance, it has 1 white temple to its right-behind in view 2.\nIn view 3, locate all target instances 'white temple' that lie in the right-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[704.0, 188.0, 1000.0, 618.0]], "image_paths": ["images/img_000832.png", "images/img_000820.png", "images/img_000829.png"], "supporting": {"draw_boxes": {"images/img_000820.png": []}}} +{"q_idx": "detect_001404", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('bench made of black metal'). For every instance, it has 1 beach ball to its left-behind in view 3.\nIn view 1, locate all target instances 'beach ball' that lie in the left-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[701.0, 717.0, 723.0, 743.0], [769.0, 642.0, 789.0, 667.0], [861.0, 542.0, 880.0, 565.0]], "image_paths": ["images/img_000635.png", "images/img_000634.png", "images/img_000628.png"], "supporting": {"draw_boxes": {"images/img_000628.png": []}}} +{"q_idx": "detect_001405", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('airplane with white and green livery'). For every instance, it has 1 airplane with yellow livery to its behind in view 2.\nIn view 3, locate all target instances 'airplane with white and green livery' that lie in the left-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[429.0, 650.0, 635.0, 775.0]], "image_paths": ["images/img_000211.png", "images/img_000205.png", "images/img_000212.png"], "supporting": {"draw_boxes": {"images/img_000205.png": []}}} +{"q_idx": "detect_001406", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('lamp'). For every instance, it has 2 lamps to its front, and 1 forklift to its right-front in view 2.\nIn view 3, locate all target instances 'lamp' that lie in the left-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[564.0, 793.0, 613.0, 870.0], [619.0, 735.0, 664.0, 808.0]], "image_paths": ["images/img_000476.png", "images/img_000466.png", "images/img_000473.png"], "supporting": {"draw_boxes": {"images/img_000466.png": []}}} +{"q_idx": "detect_001407", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('woman wearing underwear'). For every instance, it has 1 vendor cart with wooden roof to its right-front in view 2.\nIn view 1, locate all target instances 'woman wearing underwear' that lie in the left-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[385.0, 405.0, 430.0, 568.0]], "image_paths": ["images/img_001088.png", "images/img_001081.png"], "supporting": {"draw_boxes": {"images/img_001081.png": []}}} +{"q_idx": "detect_001408", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('lamp'). For every instance, it has 1 lamp to its left, and 1 worker wearing red hat and white shirt and 1 lamp to its right in view 2.\nIn view 3, locate all target instances 'lamp' that lie in the direct-left direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[234.0, 343.0, 259.0, 388.0]], "image_paths": ["images/img_000473.png", "images/img_000466.png", "images/img_000472.png"], "supporting": {"draw_boxes": {"images/img_000466.png": []}}} +{"q_idx": "detect_001409", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('bench made of blue painted metal'). For every instance, it has 1 tennis ball to its right-behind, and 1 bench made of blue painted metal to its right in view 1.\nIn view 2, locate all target instances 'tennis ball' that lie in the direct-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[728.0, 758.0, 741.0, 777.0]], "image_paths": ["images/img_000611.png", "images/img_000619.png", "images/img_000620.png"], "supporting": {"draw_boxes": {"images/img_000611.png": []}}} +{"q_idx": "detect_001410", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('striped scarf snowman'). For every instance, it has 1 coupe to its left-behind in view 3.\nIn view 2, locate all target instances 'coupe' that lie in the left-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[647.0, 87.0, 841.0, 310.0]], "image_paths": ["images/img_000733.png", "images/img_000732.png", "images/img_000722.png"], "supporting": {"draw_boxes": {"images/img_000722.png": []}}} +{"q_idx": "detect_001411", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('man wearing brown coat'). For every instance, it has 1 man wearing black coat to its behind in view 2.\nIn view 1, locate all target instances 'man wearing brown coat' that lie in the right-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[522.0, 435.0, 583.0, 668.0]], "image_paths": ["images/img_000043.png", "images/img_000035.png"], "supporting": {"draw_boxes": {"images/img_000035.png": []}}} +{"q_idx": "detect_001412", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('stone bench without armrests'). For every instance, it has 2 basketballs to its left-behind in view 2.\nIn view 1, locate all target instances 'basketball' that lie in the left-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[188.0, 467.0, 207.0, 493.0], [220.0, 377.0, 237.0, 400.0], [224.0, 365.0, 241.0, 388.0]], "image_paths": ["images/img_000587.png", "images/img_000581.png"], "supporting": {"draw_boxes": {"images/img_000581.png": []}}} +{"q_idx": "detect_001413", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('green wooden bench with armrests'). For every instance, it has 1 basketball to its right-front in view 3.\nIn view 2, locate all target instances 'basketball' that lie in the left-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[831.0, 602.0, 855.0, 632.0]], "image_paths": ["images/img_000636.png", "images/img_000634.png", "images/img_000628.png"], "supporting": {"draw_boxes": {"images/img_000628.png": []}}} +{"q_idx": "detect_001414", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('F-16 fighter jet'). For every instance, it has 1 F-16 fighter jet to its behind, and 1 F-22 fighter jet with gray color to its front in view 2.\nIn view 1, locate all target instances 'F-22 fighter jet with gray color' that lie in the right-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[226.0, 600.0, 421.0, 720.0]], "image_paths": ["images/img_000909.png", "images/img_000903.png"], "supporting": {"draw_boxes": {"images/img_000903.png": []}}} +{"q_idx": "detect_001415", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('surfboard'). For every instance, it has 1 swim ring to its left-front in view 2.\nIn view 1, locate all target instances 'swim ring' that lie in the left-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[603.0, 585.0, 657.0, 647.0]], "image_paths": ["images/img_000670.png", "images/img_000660.png", "images/img_000665.png"], "supporting": {"draw_boxes": {"images/img_000660.png": []}}} +{"q_idx": "detect_001416", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('man wearing black coat'). For every instance, it has 1 man wearing gray windbreaker to its left in view 1.\nIn view 2, locate all target instances 'man wearing black coat' that lie in the left-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[330.0, 282.0, 379.0, 493.0]], "image_paths": ["images/img_000017.png", "images/img_000025.png"], "supporting": {"draw_boxes": {"images/img_000017.png": []}}} +{"q_idx": "detect_001417", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('green wooden bench with armrests'). For every instance, it has 1 bench made of blue painted metal to its right in view 3.\nIn view 2, locate all target instances 'bench made of blue painted metal' that lie in the direct-right direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[440.0, 227.0, 585.0, 375.0]], "image_paths": ["images/img_000570.png", "images/img_000569.png", "images/img_000560.png"], "supporting": {"draw_boxes": {"images/img_000560.png": []}}} +{"q_idx": "detect_001418", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('woman wearing black hat'). For every instance, it has 1 woman wearing black hat to its left, and 1 food stall with milk branding to its left-front in view 2.\nIn view 1, locate all target instances 'woman wearing black hat' that lie in the direct-left direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[586.0, 415.0, 616.0, 555.0]], "image_paths": ["images/img_001087.png", "images/img_001077.png", "images/img_001089.png"], "supporting": {"draw_boxes": {"images/img_001077.png": []}}} +{"q_idx": "detect_001419", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('beige upright tombstone with cross'). For every instance, it has 1 tombstone with pointed top to its front in view 2.\nIn view 1, locate all target instances 'tombstone with pointed top' that lie in the direct-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[301.0, 252.0, 339.0, 360.0]], "image_paths": ["images/img_001132.png", "images/img_001126.png"], "supporting": {"draw_boxes": {"images/img_001126.png": []}}} +{"q_idx": "detect_001420", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('F-16 fighter jet'). For every instance, it has 1 airplane with red livery to its left-front in view 1.\nIn view 2, locate all target instances 'airplane with red livery' that lie in the left-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[284.0, 372.0, 419.0, 463.0]], "image_paths": ["images/img_000308.png", "images/img_000313.png", "images/img_000318.png"], "supporting": {"draw_boxes": {"images/img_000308.png": []}}} +{"q_idx": "detect_001421", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('woman in white clothing'). For every instance, it has 1 food stall with red and white striped roof to its right-behind in view 2.\nIn view 1, locate all target instances 'food stall with red and white striped roof' that lie in the right-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[755.0, 340.0, 895.0, 557.0]], "image_paths": ["images/img_001089.png", "images/img_001080.png", "images/img_001088.png"], "supporting": {"draw_boxes": {"images/img_001080.png": []}}} +{"q_idx": "detect_001422", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('mirage fighter jet'). For every instance, it has 1 stylized F22 fighter jet with camoflague color to its front in view 1.\nIn view 2, locate all target instances 'stylized F22 fighter jet with camoflague color' that lie in the direct-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[505.0, 358.0, 565.0, 402.0]], "image_paths": ["images/img_000296.png", "images/img_000303.png"], "supporting": {"draw_boxes": {"images/img_000296.png": []}}} +{"q_idx": "detect_001423", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('person with tattoos'). For every instance, it has 1 pink flamingo swim ring to its left in view 2.\nIn view 1, locate all target instances 'pink flamingo swim ring' that lie in the direct-left direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[352.0, 397.0, 404.0, 450.0]], "image_paths": ["images/img_000718.png", "images/img_000712.png", "images/img_000719.png"], "supporting": {"draw_boxes": {"images/img_000712.png": []}}} +{"q_idx": "detect_001424", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('woman in white clothing'). For every instance, it has 1 woman wearing black dress to its front in view 3.\nIn view 2, locate all target instances 'woman wearing black dress' that lie in the direct-left direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[438.0, 452.0, 479.0, 607.0]], "image_paths": ["images/img_001124.png", "images/img_001121.png", "images/img_001111.png"], "supporting": {"draw_boxes": {"images/img_001111.png": []}}} +{"q_idx": "detect_001425", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('black upright tombstone with cross'). For every instance, it has 1 temple to its right-front in view 1.\nIn view 2, locate all target instances 'temple' that lie in the right-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[114.0, 0.0, 439.0, 432.0]], "image_paths": ["images/img_000871.png", "images/img_000879.png"], "supporting": {"draw_boxes": {"images/img_000871.png": []}}} +{"q_idx": "detect_001426", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('airplane with red livery'). For every instance, it has 1 airplane with light blue livery to its behind in view 1.\nIn view 2, locate all target instances 'airplane with light blue livery' that lie in the direct-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[679.0, 430.0, 831.0, 483.0]], "image_paths": ["images/img_000308.png", "images/img_000315.png"], "supporting": {"draw_boxes": {"images/img_000308.png": []}}} +{"q_idx": "detect_001427", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('desert camouflage helicopter'). For every instance, it has 1 olive green helicopter to its right in view 2.\nIn view 3, locate all target instances 'olive green helicopter' that lie in the right-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[537.0, 620.0, 698.0, 688.0]], "image_paths": ["images/img_001023.png", "images/img_001018.png", "images/img_001024.png"], "supporting": {"draw_boxes": {"images/img_001018.png": []}}} +{"q_idx": "detect_001428", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('man wearing red jacket'). For every instance, it has 1 man wearing black coat to its front, and 1 man wearing red jacket to its left-behind in view 2.\nIn view 3, locate all target instances 'man wearing red jacket' that lie in the left-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[409.0, 268.0, 461.0, 458.0]], "image_paths": ["images/img_000024.png", "images/img_000016.png", "images/img_000021.png"], "supporting": {"draw_boxes": {"images/img_000016.png": []}}} +{"q_idx": "detect_001429", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('swim ring'). For every instance, it has 1 surfboard to its right-behind in view 2.\nIn view 1, locate all target instances 'surfboard' that lie in the right-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[355.0, 465.0, 388.0, 707.0], [328.0, 235.0, 381.0, 468.0]], "image_paths": ["images/img_000667.png", "images/img_000660.png", "images/img_000671.png"], "supporting": {"draw_boxes": {"images/img_000660.png": []}}} +{"q_idx": "detect_001430", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('tombstone without cross'). For every instance, it has 1 black flat tombstone with cross to its right in view 3.\nIn view 1, locate all target instances 'black flat tombstone with cross' that lie in the direct-right direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[403.0, 277.0, 526.0, 368.0]], "image_paths": ["images/img_001151.png", "images/img_001148.png", "images/img_001144.png"], "supporting": {"draw_boxes": {"images/img_001144.png": []}}} +{"q_idx": "detect_001431", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('woman in gray clothing'). For every instance, it has 1 woman in white clothing to its left-behind in view 3.\nIn view 1, locate all target instances 'woman in white clothing' that lie in the left-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[516.0, 342.0, 551.0, 465.0]], "image_paths": ["images/img_001056.png", "images/img_001054.png", "images/img_001049.png"], "supporting": {"draw_boxes": {"images/img_001049.png": []}}} +{"q_idx": "detect_001432", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('person with black hat'). For every instance, it has 1 person saluting to its left in view 2.\nIn view 1, locate all target instances 'person saluting' that lie in the direct-left direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[549.0, 280.0, 605.0, 413.0]], "image_paths": ["images/img_000688.png", "images/img_000679.png", "images/img_000684.png"], "supporting": {"draw_boxes": {"images/img_000679.png": []}}} +{"q_idx": "detect_001433", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('stylized F22 fighter jet with white and red color'). For every instance, it has 1 black and white striped hot air balloon to its right-behind in view 1.\nIn view 2, locate all target instances 'stylized F22 fighter jet with white and red color' that lie in the direct-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[632.0, 365.0, 771.0, 467.0]], "image_paths": ["images/img_000919.png", "images/img_000927.png", "images/img_000925.png"], "supporting": {"draw_boxes": {"images/img_000919.png": []}}} +{"q_idx": "detect_001434", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('man wearing gray hat'). For every instance, it has 2 man wearing santa costumes to its behind in view 2.\nIn view 1, locate all target instances 'man wearing santa costume' that lie in the direct-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[464.0, 240.0, 525.0, 437.0], [516.0, 267.0, 581.0, 462.0]], "image_paths": ["images/img_000055.png", "images/img_000047.png"], "supporting": {"draw_boxes": {"images/img_000047.png": []}}} +{"q_idx": "detect_001435", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('F-16 fighter jet'). For every instance, it has 1 airplane with white and orange livery to its left-front in view 1.\nIn view 2, locate all target instances 'airplane with white and orange livery' that lie in the direct-left direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[330.0, 550.0, 489.0, 632.0]], "image_paths": ["images/img_000293.png", "images/img_000300.png", "images/img_000298.png"], "supporting": {"draw_boxes": {"images/img_000293.png": []}}} +{"q_idx": "detect_001436", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('harrier jumpjet fighter jet'). For every instance, it has 1 mirage fighter jet to its behind in view 3.\nIn view 2, locate all target instances 'mirage fighter jet' that lie in the right-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[113.0, 547.0, 188.0, 622.0]], "image_paths": ["images/img_000329.png", "images/img_000334.png", "images/img_000326.png"], "supporting": {"draw_boxes": {"images/img_000326.png": []}}} +{"q_idx": "detect_001437", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('airplane with white and green livery'). For every instance, it has 2 airplane with red liveries to its behind in view 1.\nIn view 3, locate all target instances 'airplane with red livery' that lie in the left-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[427.0, 648.0, 634.0, 773.0]], "image_paths": ["images/img_000190.png", "images/img_000198.png", "images/img_000197.png"], "supporting": {"draw_boxes": {"images/img_000190.png": []}}} +{"q_idx": "detect_001438", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('snowman with black hat and stick arms'). For every instance, it has 1 man wearing green hood to its right-behind, and 1 man wearing green hood to its behind in view 1.\nIn view 3, locate all target instances 'man wearing green hood' that lie in the right-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[672.0, 333.0, 704.0, 435.0]], "image_paths": ["images/img_000743.png", "images/img_000747.png", "images/img_000753.png"], "supporting": {"draw_boxes": {"images/img_000743.png": []}}} +{"q_idx": "detect_001439", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('man wearing brown coat'). For every instance, it has 1 man wearing santa costume to its front in view 1.\nIn view 3, locate all target instances 'man wearing santa costume' that lie in the right-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[621.0, 365.0, 684.0, 597.0], [372.0, 418.0, 441.0, 663.0]], "image_paths": ["images/img_000064.png", "images/img_000072.png", "images/img_000071.png"], "supporting": {"draw_boxes": {"images/img_000064.png": []}}} +{"q_idx": "detect_001440", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('man wearing brown coat'). For every instance, it has 1 man wearing black coat to its behind in view 1.\nIn view 3, locate all target instances 'man wearing brown coat' that lie in the right-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[646.0, 323.0, 694.0, 515.0]], "image_paths": ["images/img_000035.png", "images/img_000037.png", "images/img_000041.png"], "supporting": {"draw_boxes": {"images/img_000035.png": []}}} +{"q_idx": "detect_001441", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('tombstone with shoulder top'). For every instance, it has 1 tombstone with pointed top to its front in view 3.\nIn view 1, locate all target instances 'tombstone with pointed top' that lie in the left-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[264.0, 495.0, 338.0, 603.0]], "image_paths": ["images/img_001180.png", "images/img_001181.png", "images/img_001172.png"], "supporting": {"draw_boxes": {"images/img_001172.png": []}}} +{"q_idx": "detect_001442", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('concrete mixer'). For every instance, it has 2 workers to its behind in view 2.\nIn view 1, locate all target instances 'worker' that lie in the direct-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[450.0, 263.0, 470.0, 325.0], [347.0, 267.0, 370.0, 325.0]], "image_paths": ["images/img_000444.png", "images/img_000436.png"], "supporting": {"draw_boxes": {"images/img_000436.png": []}}} +{"q_idx": "detect_001443", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('bench made of natural wood'). For every instance, it has 2 soccer balls to its right-front in view 2.\nIn view 3, locate all target instances 'soccer ball' that lie in the right-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[647.0, 765.0, 671.0, 797.0]], "image_paths": ["images/img_000606.png", "images/img_000594.png", "images/img_000601.png"], "supporting": {"draw_boxes": {"images/img_000594.png": []}}} +{"q_idx": "detect_001444", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('tombstone with shoulder top'). For every instance, it has 1 flat tombstone to its behind in view 1.\nIn view 2, locate all target instances 'tombstone with shoulder top' that lie in the left-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[541.0, 667.0, 570.0, 777.0]], "image_paths": ["images/img_001172.png", "images/img_001182.png"], "supporting": {"draw_boxes": {"images/img_001172.png": []}}} +{"q_idx": "detect_001445", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('man wearing a green hoodie and orange shoes'). For every instance, it has 1 colonnade with open top to its right-behind in view 2.\nIn view 1, locate all target instances 'colonnade with open top' that lie in the right-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[0.0, 5.0, 266.0, 860.0]], "image_paths": ["images/img_000876.png", "images/img_000870.png"], "supporting": {"draw_boxes": {"images/img_000870.png": []}}} +{"q_idx": "detect_001446", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('stylized F22 fighter jet with grey and cyan color'). For every instance, it has 1 F-22 fighter jet with gray color to its left in view 2.\nIn view 1, locate all target instances 'F-22 fighter jet with gray color' that lie in the direct-left direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[124.0, 488.0, 311.0, 575.0]], "image_paths": ["images/img_000926.png", "images/img_000920.png", "images/img_000924.png"], "supporting": {"draw_boxes": {"images/img_000920.png": []}}} +{"q_idx": "detect_001447", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('person with black hat'). For every instance, it has 1 lifeguard tower to its left-behind in view 3.\nIn view 1, locate all target instances 'lifeguard tower' that lie in the left-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[65.0, 2.0, 635.0, 242.0]], "image_paths": ["images/img_000703.png", "images/img_000698.png", "images/img_000695.png"], "supporting": {"draw_boxes": {"images/img_000695.png": []}}} +{"q_idx": "detect_001448", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('black tombstone'). For every instance, it has 1 white temple to its left-behind in view 3.\nIn view 1, locate all target instances 'white temple' that lie in the left-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[555.0, 0.0, 879.0, 433.0]], "image_paths": ["images/img_000814.png", "images/img_000818.png", "images/img_000805.png"], "supporting": {"draw_boxes": {"images/img_000805.png": []}}} +{"q_idx": "detect_001449", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('airplane with red livery'). For every instance, it has 1 stylized F22 fighter jet with grey and cyan color to its left-front in view 1.\nIn view 3, locate all target instances 'airplane with red livery' that lie in the direct-right direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[650.0, 480.0, 814.0, 558.0]], "image_paths": ["images/img_000267.png", "images/img_000276.png", "images/img_000273.png"], "supporting": {"draw_boxes": {"images/img_000267.png": []}}} +{"q_idx": "detect_001450", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('beige sand helicopter'). For every instance, it has 1 green camouflage helicopter to its left in view 2.\nIn view 1, locate all target instances 'green camouflage helicopter' that lie in the left-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[198.0, 405.0, 307.0, 463.0]], "image_paths": ["images/img_001007.png", "images/img_000997.png"], "supporting": {"draw_boxes": {"images/img_000997.png": []}}} +{"q_idx": "detect_001451", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('man wearing black coat'). For every instance, it has 1 man wearing red jacket to its behind in view 2.\nIn view 1, locate all target instances 'man wearing red jacket' that lie in the right-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[629.0, 335.0, 686.0, 555.0]], "image_paths": ["images/img_000065.png", "images/img_000058.png"], "supporting": {"draw_boxes": {"images/img_000058.png": []}}} +{"q_idx": "detect_001452", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('man wearing brown coat'). For every instance, it has 2 man wearing red jackets to its front in view 2.\nIn view 3, locate all target instances 'man wearing red jacket' that lie in the direct-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[540.0, 272.0, 591.0, 460.0], [405.0, 270.0, 463.0, 457.0]], "image_paths": ["images/img_000010.png", "images/img_000005.png", "images/img_000013.png"], "supporting": {"draw_boxes": {"images/img_000005.png": []}}} +{"q_idx": "detect_001453", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('tombstone with shoulder top'). For every instance, it has 1 beige upright tombstone with cross to its right-behind in view 1.\nIn view 3, locate all target instances 'beige upright tombstone with cross' that lie in the right-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[529.0, 198.0, 559.0, 325.0]], "image_paths": ["images/img_001126.png", "images/img_001135.png", "images/img_001134.png"], "supporting": {"draw_boxes": {"images/img_001126.png": []}}} +{"q_idx": "detect_001454", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('person with black hat'). For every instance, it has 1 person saluting to its right, and 1 person saluting to its left in view 1.\nIn view 3, locate all target instances 'person saluting' that lie in the direct-right direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[261.0, 372.0, 314.0, 502.0]], "image_paths": ["images/img_000695.png", "images/img_000704.png", "images/img_000702.png"], "supporting": {"draw_boxes": {"images/img_000695.png": []}}} +{"q_idx": "detect_001455", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('surfboard'). For every instance, it has 1 swim ring to its left-front in view 2.\nIn view 1, locate all target instances 'swim ring' that lie in the left-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[357.0, 420.0, 407.0, 490.0]], "image_paths": ["images/img_000666.png", "images/img_000660.png"], "supporting": {"draw_boxes": {"images/img_000660.png": []}}} +{"q_idx": "detect_001456", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('F-16 fighter jet'). For every instance, it has 1 airplane with yellow livery to its right-front in view 3.\nIn view 1, locate all target instances 'airplane with yellow livery' that lie in the right-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[583.0, 372.0, 718.0, 463.0]], "image_paths": ["images/img_000319.png", "images/img_000317.png", "images/img_000311.png"], "supporting": {"draw_boxes": {"images/img_000311.png": []}}} +{"q_idx": "detect_001457", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('red gloves snowman'). For every instance, it has 1 man wearing santa costume to its left-front, 1 man wearing santa costume to its right-front, and 1 red gloves snowman to its right in view 2.\nIn view 1, locate all target instances 'man wearing santa costume' that lie in the left-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[278.0, 462.0, 320.0, 545.0]], "image_paths": ["images/img_000783.png", "images/img_000777.png"], "supporting": {"draw_boxes": {"images/img_000777.png": []}}} +{"q_idx": "detect_001458", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('person with tattoos'). For every instance, it has 1 beach ball to its right-front in view 3.\nIn view 2, locate all target instances 'beach ball' that lie in the right-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[396.0, 352.0, 430.0, 397.0]], "image_paths": ["images/img_000716.png", "images/img_000720.png", "images/img_000712.png"], "supporting": {"draw_boxes": {"images/img_000712.png": []}}} +{"q_idx": "detect_001459", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('man wearing a green hoodie and orange shoes'). For every instance, it has 1 colonnade with open top to its right-front in view 3.\nIn view 2, locate all target instances 'colonnade with open top' that lie in the right-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[0.0, 0.0, 466.0, 475.0]], "image_paths": ["images/img_000812.png", "images/img_000814.png", "images/img_000806.png"], "supporting": {"draw_boxes": {"images/img_000806.png": []}}} +{"q_idx": "detect_001460", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('man wearing red jacket'). For every instance, it has 1 man wearing black coat to its front in view 2.\nIn view 1, locate all target instances 'man wearing red jacket' that lie in the left-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[620.0, 352.0, 679.0, 580.0]], "image_paths": ["images/img_000024.png", "images/img_000017.png"], "supporting": {"draw_boxes": {"images/img_000017.png": []}}} +{"q_idx": "detect_001461", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('mirage fighter jet'). For every instance, it has 1 white and green striped hot air balloon to its left-behind in view 2.\nIn view 3, locate all target instances 'mirage fighter jet' that lie in the direct-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[188.0, 438.0, 285.0, 583.0]], "image_paths": ["images/img_000938.png", "images/img_000936.png", "images/img_000944.png"], "supporting": {"draw_boxes": {"images/img_000936.png": []}}} +{"q_idx": "detect_001462", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('F-22 fighter jet with gray color'). For every instance, it has 1 mirage fighter jet to its left in view 1.\nIn view 3, locate all target instances 'mirage fighter jet' that lie in the direct-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[781.0, 453.0, 940.0, 522.0]], "image_paths": ["images/img_000886.png", "images/img_000891.png", "images/img_000892.png"], "supporting": {"draw_boxes": {"images/img_000886.png": []}}} +{"q_idx": "detect_001463", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('bench made of stone'). For every instance, it has 2 tennis balls to its right-behind in view 2.\nIn view 1, locate all target instances 'tennis ball' that lie in the right-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[795.0, 550.0, 807.0, 567.0]], "image_paths": ["images/img_000587.png", "images/img_000581.png"], "supporting": {"draw_boxes": {"images/img_000581.png": []}}} +{"q_idx": "detect_001464", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('person saluting'). For every instance, it has 1 person saluting to its right, and 1 lifeguard tower to its behind in view 3.\nIn view 1, locate all target instances 'lifeguard tower' that lie in the direct-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[3.0, 0.0, 409.0, 490.0]], "image_paths": ["images/img_000702.png", "images/img_000700.png", "images/img_000695.png"], "supporting": {"draw_boxes": {"images/img_000695.png": []}}} +{"q_idx": "detect_001465", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('white upright tombstone with cross'). For every instance, it has 1 tombstone with pointed top and 1 white upright tombstone with cross to its left in view 2.\nIn view 1, locate all target instances 'tombstone with pointed top' that lie in the direct-left direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[743.0, 425.0, 814.0, 555.0]], "image_paths": ["images/img_001139.png", "images/img_001130.png"], "supporting": {"draw_boxes": {"images/img_001130.png": []}}} +{"q_idx": "detect_001466", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('man wearing santa costume'). For every instance, it has 1 man wearing santa costume to its front, and 1 taxi to its left-front in view 3.\nIn view 1, locate all target instances 'taxi' that lie in the left-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[0.0, 407.0, 188.0, 538.0]], "image_paths": ["images/img_000801.png", "images/img_000797.png", "images/img_000794.png"], "supporting": {"draw_boxes": {"images/img_000794.png": []}}} +{"q_idx": "detect_001467", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('man in crouching posture'). For every instance, it has 1 man in standing posture to its left in view 1.\nIn view 3, locate all target instances 'man in standing posture' that lie in the direct-left direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[438.0, 318.0, 469.0, 408.0]], "image_paths": ["images/img_000825.png", "images/img_000830.png", "images/img_000832.png"], "supporting": {"draw_boxes": {"images/img_000825.png": []}}} +{"q_idx": "detect_001468", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('worker'). For every instance, it has 1 worker to its behind, and 1 truck with white and black cab to its left-front in view 1.\nIn view 3, locate all target instances 'truck with white and black cab' that lie in the left-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[150.0, 147.0, 359.0, 352.0]], "image_paths": ["images/img_000453.png", "images/img_000462.png", "images/img_000460.png"], "supporting": {"draw_boxes": {"images/img_000453.png": []}}} +{"q_idx": "detect_001469", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('man wearing santa costume'). For every instance, it has 1 man wearing red jacket to its left in view 1.\nIn view 2, locate all target instances 'man wearing santa costume' that lie in the direct-right direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[278.0, 462.0, 320.0, 545.0]], "image_paths": ["images/img_000776.png", "images/img_000783.png", "images/img_000786.png"], "supporting": {"draw_boxes": {"images/img_000776.png": []}}} +{"q_idx": "detect_001470", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('bench made of black metal'). For every instance, it has 1 tennis ball to its right-behind, 1 tennis ball to its behind, and 2 bench made of black metals to its right in view 3.\nIn view 1, locate all target instances 'bench made of black metal' that lie in the direct-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[637.0, 265.0, 721.0, 422.0]], "image_paths": ["images/img_000638.png", "images/img_000637.png", "images/img_000632.png"], "supporting": {"draw_boxes": {"images/img_000632.png": []}}} +{"q_idx": "detect_001471", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('surfboard'). For every instance, it has 1 surfboard to its right, and 1 beach cabin to its right-front in view 1.\nIn view 3, locate all target instances 'surfboard' that lie in the direct-right direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[578.0, 197.0, 604.0, 427.0]], "image_paths": ["images/img_000663.png", "images/img_000672.png", "images/img_000671.png"], "supporting": {"draw_boxes": {"images/img_000663.png": []}}} +{"q_idx": "detect_001472", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('airplane with yellow livery'). For every instance, it has 1 airplane with white and orange livery to its left-front, and 1 airplane with yellow livery to its front in view 2.\nIn view 1, locate all target instances 'airplane with yellow livery' that lie in the left-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[691.0, 468.0, 879.0, 553.0]], "image_paths": ["images/img_000254.png", "images/img_000248.png", "images/img_000261.png"], "supporting": {"draw_boxes": {"images/img_000248.png": []}}} +{"q_idx": "detect_001473", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('airplane with yellow livery'). For every instance, it has 1 airplane with white and orange livery to its behind in view 1.\nIn view 3, locate all target instances 'airplane with white and orange livery' that lie in the direct-right direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[532.0, 390.0, 647.0, 433.0]], "image_paths": ["images/img_000294.png", "images/img_000304.png", "images/img_000300.png"], "supporting": {"draw_boxes": {"images/img_000294.png": []}}} +{"q_idx": "detect_001474", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('mirage fighter jet'). For every instance, it has 1 airplane with yellow livery to its right-front in view 1.\nIn view 2, locate all target instances 'mirage fighter jet' that lie in the right-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[113.0, 547.0, 188.0, 622.0]], "image_paths": ["images/img_000326.png", "images/img_000334.png"], "supporting": {"draw_boxes": {"images/img_000326.png": []}}} +{"q_idx": "detect_001475", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('airplane with yellow livery'). For every instance, it has 1 airplane with white and green livery to its left, and 1 airplane with white and green livery to its right in view 3.\nIn view 1, locate all target instances 'airplane with white and green livery' that lie in the left-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[46.0, 453.0, 214.0, 508.0]], "image_paths": ["images/img_000214.png", "images/img_000211.png", "images/img_000207.png"], "supporting": {"draw_boxes": {"images/img_000207.png": []}}} +{"q_idx": "detect_001476", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('fan'). For every instance, it has 1 truck with white and black cab to its right-behind in view 1.\nIn view 3, locate all target instances 'truck with white and black cab' that lie in the right-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[0.0, 270.0, 190.0, 512.0]], "image_paths": ["images/img_000453.png", "images/img_000458.png", "images/img_000459.png"], "supporting": {"draw_boxes": {"images/img_000453.png": []}}} +{"q_idx": "detect_001477", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('red gloves snowman'). For every instance, it has 1 man wearing gray windbreaker to its left-behind, 1 man wearing gray windbreaker to its left, and 1 man wearing gray windbreaker to its left-front in view 2.\nIn view 1, locate all target instances 'man wearing gray windbreaker' that lie in the left-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[335.0, 227.0, 375.0, 312.0]], "image_paths": ["images/img_000781.png", "images/img_000773.png"], "supporting": {"draw_boxes": {"images/img_000773.png": []}}} +{"q_idx": "detect_001478", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('green camouflage helicopter'). For every instance, it has 2 dark grey helicopters to its right in view 3.\nIn view 2, locate all target instances 'dark grey helicopter' that lie in the direct-right direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[65.0, 590.0, 226.0, 692.0], [199.0, 403.0, 310.0, 463.0]], "image_paths": ["images/img_001036.png", "images/img_001040.png", "images/img_001035.png"], "supporting": {"draw_boxes": {"images/img_001035.png": []}}} +{"q_idx": "detect_001479", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('red and white helicopter'). For every instance, it has 1 black and white striped hot air balloon to its left-front in view 1.\nIn view 3, locate all target instances 'black and white striped hot air balloon' that lie in the left-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[10.0, 285.0, 150.0, 503.0]], "image_paths": ["images/img_000964.png", "images/img_000972.png", "images/img_000971.png"], "supporting": {"draw_boxes": {"images/img_000964.png": []}}} +{"q_idx": "detect_001480", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('stone bench without armrests'). For every instance, it has 2 tennis balls to its right-behind in view 1.\nIn view 2, locate all target instances 'tennis ball' that lie in the right-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[231.0, 392.0, 241.0, 405.0], [181.0, 420.0, 193.0, 433.0]], "image_paths": ["images/img_000581.png", "images/img_000588.png"], "supporting": {"draw_boxes": {"images/img_000581.png": []}}} +{"q_idx": "detect_001481", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('tombstone with round top'). For every instance, it has 1 man with pink pants carrying a phone and a shopping bag to its left in view 1.\nIn view 2, locate all target instances 'man with pink pants carrying a phone and a shopping bag' that lie in the direct-left direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[616.0, 463.0, 649.0, 555.0]], "image_paths": ["images/img_000868.png", "images/img_000878.png", "images/img_000883.png"], "supporting": {"draw_boxes": {"images/img_000868.png": []}}} +{"q_idx": "detect_001482", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('red and white helicopter'). For every instance, it has 1 dark grey helicopter to its behind in view 1.\nIn view 3, locate all target instances 'dark grey helicopter' that lie in the direct-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[815.0, 432.0, 935.0, 512.0]], "image_paths": ["images/img_001029.png", "images/img_001036.png", "images/img_001037.png"], "supporting": {"draw_boxes": {"images/img_001029.png": []}}} +{"q_idx": "detect_001483", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('lamp'). For every instance, it has 2 lamps to its right, and 1 forklift to its right-front in view 2.\nIn view 1, locate all target instances 'forklift' that lie in the right-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[196.0, 215.0, 302.0, 372.0]], "image_paths": ["images/img_000473.png", "images/img_000466.png", "images/img_000475.png"], "supporting": {"draw_boxes": {"images/img_000466.png": []}}} +{"q_idx": "detect_001484", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('man wearing a dark kurta and white pants'). For every instance, it has 1 man wearing a dark kurta and white pants to its behind, and 1 colonnade with rectangular top to its right-behind in view 1.\nIn view 2, locate all target instances 'colonnade with rectangular top' that lie in the right-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[544.0, 52.0, 896.0, 462.0]], "image_paths": ["images/img_000839.png", "images/img_000848.png"], "supporting": {"draw_boxes": {"images/img_000839.png": []}}} +{"q_idx": "detect_001485", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('red buttons snowman'). For every instance, it has 2 no hat snowmans to its right in view 3.\nIn view 1, locate all target instances 'no hat snowman' that lie in the direct-right direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[453.0, 252.0, 507.0, 357.0], [375.0, 325.0, 435.0, 430.0]], "image_paths": ["images/img_000763.png", "images/img_000766.png", "images/img_000756.png"], "supporting": {"draw_boxes": {"images/img_000756.png": []}}} +{"q_idx": "detect_001486", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('stone bench without armrests'). For every instance, it has 1 baseball to its right-front in view 3.\nIn view 1, locate all target instances 'baseball' that lie in the right-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[301.0, 730.0, 318.0, 752.0], [210.0, 620.0, 225.0, 640.0]], "image_paths": ["images/img_000586.png", "images/img_000589.png", "images/img_000577.png"], "supporting": {"draw_boxes": {"images/img_000577.png": []}}} +{"q_idx": "detect_001487", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('F-16 fighter jet'). For every instance, it has 1 F-22 fighter jet with gray color to its front in view 2.\nIn view 1, locate all target instances 'F-22 fighter jet with gray color' that lie in the direct-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[246.0, 352.0, 380.0, 408.0]], "image_paths": ["images/img_000891.png", "images/img_000885.png", "images/img_000892.png"], "supporting": {"draw_boxes": {"images/img_000885.png": []}}} +{"q_idx": "detect_001488", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('dark grey helicopter'). For every instance, it has 1 olive green helicopter to its right in view 1.\nIn view 3, locate all target instances 'olive green helicopter' that lie in the direct-left direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[556.0, 338.0, 584.0, 400.0]], "image_paths": ["images/img_000986.png", "images/img_000994.png", "images/img_000993.png"], "supporting": {"draw_boxes": {"images/img_000986.png": []}}} +{"q_idx": "detect_001489", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('woman wearing scarf accessory'). For every instance, it has 1 vendor cart with wooden roof to its right-front in view 2.\nIn view 3, locate all target instances 'vendor cart with wooden roof' that lie in the right-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[200.0, 192.0, 411.0, 428.0]], "image_paths": ["images/img_001058.png", "images/img_001050.png", "images/img_001056.png"], "supporting": {"draw_boxes": {"images/img_001050.png": []}}} +{"q_idx": "detect_001490", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('man wearing santa costume'). For every instance, it has 1 man wearing santa costume to its behind, and 1 man wearing gray hat to its front in view 3.\nIn view 2, locate all target instances 'man wearing santa costume' that lie in the direct-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[264.0, 327.0, 336.0, 542.0]], "image_paths": ["images/img_000052.png", "images/img_000053.png", "images/img_000046.png"], "supporting": {"draw_boxes": {"images/img_000046.png": []}}} +{"q_idx": "detect_001491", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('airplane with white and green livery'). For every instance, it has 1 airplane with light blue livery to its front in view 3.\nIn view 1, locate all target instances 'airplane with light blue livery' that lie in the direct-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[235.0, 373.0, 359.0, 418.0]], "image_paths": ["images/img_000198.png", "images/img_000200.png", "images/img_000193.png"], "supporting": {"draw_boxes": {"images/img_000193.png": []}}} +{"q_idx": "detect_001492", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('worker'). For every instance, it has 1 worker to its left, and 1 white stack of bricks to its front in view 2.\nIn view 1, locate all target instances 'white stack of bricks' that lie in the direct-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[448.0, 245.0, 546.0, 372.0]], "image_paths": ["images/img_000445.png", "images/img_000437.png"], "supporting": {"draw_boxes": {"images/img_000437.png": []}}} +{"q_idx": "detect_001493", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('three balls snowman'). For every instance, it has 2 black buttons snowmans to its left in view 1.\nIn view 2, locate all target instances 'black buttons snowman' that lie in the direct-left direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[630.0, 402.0, 716.0, 508.0], [561.0, 328.0, 640.0, 430.0]], "image_paths": ["images/img_000760.png", "images/img_000769.png"], "supporting": {"draw_boxes": {"images/img_000760.png": []}}} +{"q_idx": "detect_001494", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('white tombstone'). For every instance, it has 1 pavilion to its left-front in view 1.\nIn view 2, locate all target instances 'pavilion' that lie in the left-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[701.0, 223.0, 931.0, 570.0]], "image_paths": ["images/img_000854.png", "images/img_000863.png"], "supporting": {"draw_boxes": {"images/img_000854.png": []}}} +{"q_idx": "detect_001495", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('tombstone with shoulder top'). For every instance, it has 1 beige upright tombstone with cross to its right in view 3.\nIn view 2, locate all target instances 'beige upright tombstone with cross' that lie in the right-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[556.0, 580.0, 610.0, 715.0]], "image_paths": ["images/img_001203.png", "images/img_001197.png", "images/img_001189.png"], "supporting": {"draw_boxes": {"images/img_001189.png": []}}} +{"q_idx": "detect_001496", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('airplane with light blue livery'). For every instance, it has 1 airplane with red livery to its right-front, and 1 airplane with red livery to its front in view 3.\nIn view 1, locate all target instances 'airplane with red livery' that lie in the right-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[370.0, 373.0, 494.0, 418.0]], "image_paths": ["images/img_000241.png", "images/img_000245.png", "images/img_000235.png"], "supporting": {"draw_boxes": {"images/img_000235.png": []}}} +{"q_idx": "detect_001497", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('man wearing red jacket'). For every instance, it has 1 man wearing brown coat to its left in view 1.\nIn view 2, locate all target instances 'man wearing red jacket' that lie in the right-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[569.0, 390.0, 629.0, 632.0]], "image_paths": ["images/img_000034.png", "images/img_000043.png"], "supporting": {"draw_boxes": {"images/img_000034.png": []}}} +{"q_idx": "detect_001498", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('person with black hat'). For every instance, it has 1 person saluting to its right, and 1 person saluting to its left in view 2.\nIn view 3, locate all target instances 'person saluting' that lie in the direct-left direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[505.0, 238.0, 554.0, 373.0]], "image_paths": ["images/img_000702.png", "images/img_000695.png", "images/img_000703.png"], "supporting": {"draw_boxes": {"images/img_000695.png": []}}} +{"q_idx": "detect_001499", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('white flat tombstone with cross'). For every instance, it has 1 colonnade with circular top to its right in view 1.\nIn view 2, locate all target instances 'colonnade with circular top' that lie in the direct-right direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[0.0, 0.0, 465.0, 427.0]], "image_paths": ["images/img_000858.png", "images/img_000866.png", "images/img_000860.png"], "supporting": {"draw_boxes": {"images/img_000858.png": []}}} +{"q_idx": "detect_001500", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('man holding box'). For every instance, it has 1 man wearing a dark kurta and white pants to its left in view 1.\nIn view 2, locate all target instances 'man holding box' that lie in the right-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[364.0, 530.0, 389.0, 620.0]], "image_paths": ["images/img_000838.png", "images/img_000847.png", "images/img_000850.png"], "supporting": {"draw_boxes": {"images/img_000838.png": []}}} +{"q_idx": "detect_001501", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('airplane with red livery'). For every instance, it has 1 airplane with white and orange livery to its behind in view 2.\nIn view 1, locate all target instances 'airplane with white and orange livery' that lie in the direct-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[509.0, 360.0, 642.0, 420.0]], "image_paths": ["images/img_000239.png", "images/img_000232.png", "images/img_000242.png"], "supporting": {"draw_boxes": {"images/img_000232.png": []}}} +{"q_idx": "detect_001502", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('person with tattoos'). For every instance, it has 1 toy crab to its right-front in view 2.\nIn view 1, locate all target instances 'toy crab' that lie in the right-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[339.0, 403.0, 381.0, 452.0], [331.0, 568.0, 380.0, 630.0]], "image_paths": ["images/img_000719.png", "images/img_000712.png"], "supporting": {"draw_boxes": {"images/img_000712.png": []}}} +{"q_idx": "detect_001503", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('F-16 fighter jet'). For every instance, it has 1 mirage fighter jet to its behind, and 1 F-16 fighter jet to its left-behind in view 2.\nIn view 3, locate all target instances 'mirage fighter jet' that lie in the left-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[152.0, 463.0, 318.0, 582.0]], "image_paths": ["images/img_000895.png", "images/img_000890.png", "images/img_000898.png"], "supporting": {"draw_boxes": {"images/img_000890.png": []}}} +{"q_idx": "detect_001504", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('airplane with white and orange livery'). For every instance, it has 1 airplane with red livery to its front in view 1.\nIn view 2, locate all target instances 'airplane with white and orange livery' that lie in the left-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[636.0, 433.0, 792.0, 563.0]], "image_paths": ["images/img_000278.png", "images/img_000283.png", "images/img_000287.png"], "supporting": {"draw_boxes": {"images/img_000278.png": []}}} +{"q_idx": "detect_001505", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('natural wooden bench with armrests'). For every instance, it has 4 basketballs to its left-front, and 1 natural wooden bench with armrests to its left in view 1.\nIn view 2, locate all target instances 'basketball' that lie in the left-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[828.0, 615.0, 851.0, 645.0], [835.0, 635.0, 859.0, 667.0]], "image_paths": ["images/img_000598.png", "images/img_000604.png"], "supporting": {"draw_boxes": {"images/img_000598.png": []}}} +{"q_idx": "detect_001506", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('stylized F22 fighter jet with grey and cyan color'). For every instance, it has 1 mirage fighter jet to its front in view 1.\nIn view 2, locate all target instances 'mirage fighter jet' that lie in the direct-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[502.0, 363.0, 573.0, 393.0]], "image_paths": ["images/img_000320.png", "images/img_000329.png"], "supporting": {"draw_boxes": {"images/img_000320.png": []}}} +{"q_idx": "detect_001507", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('man wearing santa costume'). For every instance, it has 1 man wearing santa costume to its left, and 1 striped scarf snowman to its right-behind in view 2.\nIn view 1, locate all target instances 'man wearing santa costume' that lie in the direct-left direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[562.0, 182.0, 601.0, 293.0]], "image_paths": ["images/img_000736.png", "images/img_000728.png"], "supporting": {"draw_boxes": {"images/img_000728.png": []}}} +{"q_idx": "detect_001508", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('olive green helicopter'). For every instance, it has 1 red and black striped hot air balloon to its right-front in view 1.\nIn view 2, locate all target instances 'olive green helicopter' that lie in the direct-left direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[556.0, 338.0, 584.0, 400.0]], "image_paths": ["images/img_000986.png", "images/img_000993.png", "images/img_000992.png"], "supporting": {"draw_boxes": {"images/img_000986.png": []}}} +{"q_idx": "detect_001509", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('red buttons snowman'). For every instance, it has 1 man wearing santa costume to its right-behind, and 1 red buttons snowman to its right in view 2.\nIn view 1, locate all target instances 'red buttons snowman' that lie in the direct-right direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[350.0, 578.0, 426.0, 662.0]], "image_paths": ["images/img_000797.png", "images/img_000791.png"], "supporting": {"draw_boxes": {"images/img_000791.png": []}}} +{"q_idx": "detect_001510", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('black upright tombstone with cross'). For every instance, it has 1 white upright tombstone with cross and 1 black upright tombstone with cross to its left in view 3.\nIn view 1, locate all target instances 'black upright tombstone with cross' that lie in the direct-left direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[537.0, 223.0, 565.0, 343.0]], "image_paths": ["images/img_001167.png", "images/img_001166.png", "images/img_001160.png"], "supporting": {"draw_boxes": {"images/img_001160.png": []}}} +{"q_idx": "detect_001511", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('olive green helicopter'). For every instance, it has 2 beige sand helicopters to its right, and 1 beige sand helicopter to its left in view 1.\nIn view 3, locate all target instances 'beige sand helicopter' that lie in the direct-right direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[660.0, 345.0, 755.0, 408.0], [527.0, 345.0, 614.0, 408.0]], "image_paths": ["images/img_000965.png", "images/img_000976.png", "images/img_000973.png"], "supporting": {"draw_boxes": {"images/img_000965.png": []}}} +{"q_idx": "detect_001512", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('toy crab'). For every instance, it has 1 beach ball to its left in view 2.\nIn view 3, locate all target instances 'beach ball' that lie in the left-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[341.0, 547.0, 381.0, 598.0], [480.0, 690.0, 522.0, 748.0]], "image_paths": ["images/img_000719.png", "images/img_000712.png", "images/img_000720.png"], "supporting": {"draw_boxes": {"images/img_000712.png": []}}} +{"q_idx": "detect_001513", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('man wearing red jacket'). For every instance, it has 1 man wearing black coat and 1 man wearing red jacket to its front in view 3.\nIn view 1, locate all target instances 'man wearing black coat' that lie in the right-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[407.0, 427.0, 474.0, 673.0]], "image_paths": ["images/img_000012.png", "images/img_000010.png", "images/img_000005.png"], "supporting": {"draw_boxes": {"images/img_000005.png": []}}} +{"q_idx": "detect_001514", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('stylized F22 fighter jet with camoflague color'). For every instance, it has 1 black and white striped hot air balloon to its right-front in view 1.\nIn view 2, locate all target instances 'black and white striped hot air balloon' that lie in the right-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[458.0, 205.0, 541.0, 352.0]], "image_paths": ["images/img_000902.png", "images/img_000908.png"], "supporting": {"draw_boxes": {"images/img_000902.png": []}}} +{"q_idx": "detect_001515", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('man wearing gray hat'). For every instance, it has 1 bus stop with enclosure to its left-front in view 3.\nIn view 1, locate all target instances 'man wearing gray hat' that lie in the left-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[644.0, 427.0, 700.0, 652.0]], "image_paths": ["images/img_000036.png", "images/img_000039.png", "images/img_000030.png"], "supporting": {"draw_boxes": {"images/img_000030.png": []}}} +{"q_idx": "detect_001516", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('mirage fighter jet'). For every instance, it has 1 harrier jumpjet fighter jet to its right-behind in view 3.\nIn view 1, locate all target instances 'harrier jumpjet fighter jet' that lie in the right-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[359.0, 352.0, 446.0, 410.0]], "image_paths": ["images/img_000927.png", "images/img_000924.png", "images/img_000919.png"], "supporting": {"draw_boxes": {"images/img_000919.png": []}}} +{"q_idx": "detect_001517", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('man wearing green hood'). For every instance, it has 1 man wearing santa costume to its left, and 1 man wearing santa costume to its left-front in view 1.\nIn view 2, locate all target instances 'man wearing santa costume' that lie in the left-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[630.0, 367.0, 710.0, 578.0]], "image_paths": ["images/img_000047.png", "images/img_000057.png"], "supporting": {"draw_boxes": {"images/img_000047.png": []}}} +{"q_idx": "detect_001518", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('airplane with white and green livery'). For every instance, it has 1 airplane with white and orange livery to its front in view 1.\nIn view 2, locate all target instances 'airplane with white and green livery' that lie in the direct-right direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[242.0, 340.0, 362.0, 428.0], [505.0, 340.0, 624.0, 428.0]], "image_paths": ["images/img_000221.png", "images/img_000230.png"], "supporting": {"draw_boxes": {"images/img_000221.png": []}}} +{"q_idx": "detect_001519", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('red buttons snowman'). For every instance, it has 1 green sedan to its right in view 1.\nIn view 2, locate all target instances 'green sedan' that lie in the direct-right direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[180.0, 113.0, 347.0, 305.0]], "image_paths": ["images/img_000745.png", "images/img_000751.png", "images/img_000750.png"], "supporting": {"draw_boxes": {"images/img_000745.png": []}}} +{"q_idx": "detect_001520", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('airplane with light blue livery'). For every instance, it has 1 airplane with red livery to its right-behind in view 2.\nIn view 1, locate all target instances 'airplane with light blue livery' that lie in the right-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[62.0, 412.0, 206.0, 515.0]], "image_paths": ["images/img_000216.png", "images/img_000209.png"], "supporting": {"draw_boxes": {"images/img_000209.png": []}}} +{"q_idx": "detect_001521", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('white flat tombstone with cross'). For every instance, it has 1 tombstone without cross to its behind in view 2.\nIn view 1, locate all target instances 'tombstone without cross' that lie in the direct-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[319.0, 445.0, 359.0, 498.0]], "image_paths": ["images/img_000864.png", "images/img_000858.png", "images/img_000862.png"], "supporting": {"draw_boxes": {"images/img_000858.png": []}}} +{"q_idx": "detect_001522", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('beige upright tombstone with cross'). For every instance, it has 1 tombstone with shoulder top to its left in view 1.\nIn view 3, locate all target instances 'beige upright tombstone with cross' that lie in the left-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[482.0, 177.0, 524.0, 302.0]], "image_paths": ["images/img_001190.png", "images/img_001197.png", "images/img_001200.png"], "supporting": {"draw_boxes": {"images/img_001190.png": []}}} +{"q_idx": "detect_001523", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('desert camouflage helicopter'). For every instance, it has 1 olive green helicopter to its right in view 3.\nIn view 2, locate all target instances 'desert camouflage helicopter' that lie in the right-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[74.0, 577.0, 217.0, 713.0]], "image_paths": ["images/img_001024.png", "images/img_001026.png", "images/img_001018.png"], "supporting": {"draw_boxes": {"images/img_001018.png": []}}} +{"q_idx": "detect_001524", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('man wearing orange coat'). For every instance, it has 1 man wearing red jacket to its front in view 1.\nIn view 3, locate all target instances 'man wearing red jacket' that lie in the direct-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[339.0, 265.0, 395.0, 448.0]], "image_paths": ["images/img_000044.png", "images/img_000052.png", "images/img_000050.png"], "supporting": {"draw_boxes": {"images/img_000044.png": []}}} +{"q_idx": "detect_001525", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('black tombstone'). For every instance, it has 1 man wearing a green hoodie and orange shoes to its behind in view 2.\nIn view 3, locate all target instances 'man wearing a green hoodie and orange shoes' that lie in the right-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[624.0, 388.0, 651.0, 485.0]], "image_paths": ["images/img_000862.png", "images/img_000852.png", "images/img_000859.png"], "supporting": {"draw_boxes": {"images/img_000852.png": []}}} +{"q_idx": "detect_001526", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('woman in white clothing'). For every instance, it has 1 woman in gray clothing to its left-front in view 3.\nIn view 1, locate all target instances 'woman in white clothing' that lie in the left-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[616.0, 392.0, 654.0, 525.0], [573.0, 423.0, 618.0, 560.0]], "image_paths": ["images/img_001054.png", "images/img_001056.png", "images/img_001049.png"], "supporting": {"draw_boxes": {"images/img_001049.png": []}}} +{"q_idx": "detect_001527", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('airplane with white and green livery'). For every instance, it has 1 airplane with yellow livery to its left in view 1.\nIn view 2, locate all target instances 'airplane with yellow livery' that lie in the direct-left direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[366.0, 378.0, 471.0, 423.0]], "image_paths": ["images/img_000234.png", "images/img_000240.png", "images/img_000242.png"], "supporting": {"draw_boxes": {"images/img_000234.png": []}}} +{"q_idx": "detect_001528", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('worker'). For every instance, it has 1 worker to its behind, and 1 white stack of sandbags to its right-front in view 1.\nIn view 2, locate all target instances 'white stack of sandbags' that lie in the right-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[614.0, 263.0, 704.0, 415.0]], "image_paths": ["images/img_000453.png", "images/img_000461.png", "images/img_000460.png"], "supporting": {"draw_boxes": {"images/img_000453.png": []}}} +{"q_idx": "detect_001529", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('airplane with yellow livery'). For every instance, it has 1 airplane with white and green livery to its behind, and 1 airplane with white and green livery to its right-behind in view 1.\nIn view 3, locate all target instances 'airplane with white and green livery' that lie in the right-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[62.0, 412.0, 206.0, 515.0], [341.0, 348.0, 444.0, 428.0]], "image_paths": ["images/img_000222.png", "images/img_000227.png", "images/img_000229.png"], "supporting": {"draw_boxes": {"images/img_000222.png": []}}} +{"q_idx": "detect_001530", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('lamp'). For every instance, it has 2 lamps to its right, and 1 beige stone steps to its left-behind in view 3.\nIn view 1, locate all target instances 'beige stone steps' that lie in the left-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[422.0, 220.0, 595.0, 392.0]], "image_paths": ["images/img_000475.png", "images/img_000477.png", "images/img_000466.png"], "supporting": {"draw_boxes": {"images/img_000466.png": []}}} +{"q_idx": "detect_001531", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('person with black hat'). For every instance, it has 1 toy crab to its left in view 3.\nIn view 1, locate all target instances 'toy crab' that lie in the direct-left direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[399.0, 377.0, 445.0, 418.0]], "image_paths": ["images/img_000686.png", "images/img_000685.png", "images/img_000679.png"], "supporting": {"draw_boxes": {"images/img_000679.png": []}}} +{"q_idx": "detect_001532", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('bench made of black metal'). For every instance, it has 2 tennis balls to its left-behind, 1 bench made of black metal to its right, and 1 bench made of black metal to its left in view 1.\nIn view 3, locate all target instances 'tennis ball' that lie in the direct-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[210.0, 483.0, 221.0, 497.0]], "image_paths": ["images/img_000632.png", "images/img_000636.png", "images/img_000638.png"], "supporting": {"draw_boxes": {"images/img_000632.png": []}}} +{"q_idx": "detect_001533", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('tombstone with shoulder top'). For every instance, it has 1 man in blue suit and tie to its right-behind in view 3.\nIn view 1, locate all target instances 'tombstone with shoulder top' that lie in the left-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[281.0, 517.0, 355.0, 583.0]], "image_paths": ["images/img_001186.png", "images/img_001182.png", "images/img_001178.png"], "supporting": {"draw_boxes": {"images/img_001178.png": []}}} +{"q_idx": "detect_001534", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('man wearing black coat'). For every instance, it has 1 snowman with black hat and red gloves to its right-behind, and 1 snowman with black hat and red gloves to its right-front in view 2.\nIn view 1, locate all target instances 'snowman with black hat and red gloves' that lie in the right-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[532.0, 303.0, 609.0, 408.0]], "image_paths": ["images/img_000749.png", "images/img_000741.png"], "supporting": {"draw_boxes": {"images/img_000741.png": []}}} +{"q_idx": "detect_001535", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('tombstone with shoulder top'). For every instance, it has 1 tombstone with round top to its behind in view 2.\nIn view 1, locate all target instances 'tombstone with round top' that lie in the direct-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[666.0, 387.0, 705.0, 473.0]], "image_paths": ["images/img_001201.png", "images/img_001191.png", "images/img_001198.png"], "supporting": {"draw_boxes": {"images/img_001191.png": []}}} +{"q_idx": "detect_001536", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('olive green helicopter'). For every instance, it has 1 red and black striped hot air balloon to its right-front in view 1.\nIn view 2, locate all target instances 'red and black striped hot air balloon' that lie in the right-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[0.0, 348.0, 147.0, 728.0]], "image_paths": ["images/img_000986.png", "images/img_000992.png", "images/img_000994.png"], "supporting": {"draw_boxes": {"images/img_000986.png": []}}} +{"q_idx": "detect_001537", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('man wearing black coat'). For every instance, it has 1 man wearing brown coat to its front in view 2.\nIn view 1, locate all target instances 'man wearing black coat' that lie in the left-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[407.0, 427.0, 474.0, 673.0]], "image_paths": ["images/img_000012.png", "images/img_000005.png"], "supporting": {"draw_boxes": {"images/img_000005.png": []}}} +{"q_idx": "detect_001538", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('man in standing posture'). For every instance, it has 1 flat tombstone to its right in view 1.\nIn view 2, locate all target instances 'flat tombstone' that lie in the direct-right direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[339.0, 580.0, 430.0, 642.0]], "image_paths": ["images/img_000825.png", "images/img_000831.png", "images/img_000827.png"], "supporting": {"draw_boxes": {"images/img_000825.png": []}}} +{"q_idx": "detect_001539", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('black flat tombstone with cross'). For every instance, it has 1 beige tombstone to its left, and 1 beige tombstone to its right in view 2.\nIn view 1, locate all target instances 'beige tombstone' that lie in the direct-right direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[740.0, 475.0, 809.0, 557.0]], "image_paths": ["images/img_001164.png", "images/img_001156.png"], "supporting": {"draw_boxes": {"images/img_001156.png": []}}} +{"q_idx": "detect_001540", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('bench made of black metal'). For every instance, it has 1 basketball to its right-behind in view 1.\nIn view 2, locate all target instances 'basketball' that lie in the right-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[609.0, 303.0, 625.0, 323.0]], "image_paths": ["images/img_000560.png", "images/img_000568.png"], "supporting": {"draw_boxes": {"images/img_000560.png": []}}} +{"q_idx": "detect_001541", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('worker'). For every instance, it has 1 concrete mixer to its behind, and 1 worker to its front in view 2.\nIn view 1, locate all target instances 'worker' that lie in the direct-left direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[693.0, 648.0, 726.0, 743.0]], "image_paths": ["images/img_000443.png", "images/img_000436.png"], "supporting": {"draw_boxes": {"images/img_000436.png": []}}} +{"q_idx": "detect_001542", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('airplane with red livery'). For every instance, it has 1 mirage fighter jet to its left-behind in view 3.\nIn view 1, locate all target instances 'airplane with red livery' that lie in the right-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[289.0, 498.0, 444.0, 633.0]], "image_paths": ["images/img_000334.png", "images/img_000331.png", "images/img_000325.png"], "supporting": {"draw_boxes": {"images/img_000325.png": []}}} +{"q_idx": "detect_001543", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('man in sitting posture'). For every instance, it has 1 colonnade with open top to its left in view 2.\nIn view 1, locate all target instances 'colonnade with open top' that lie in the direct-left direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[0.0, 0.0, 274.0, 857.0]], "image_paths": ["images/img_000813.png", "images/img_000807.png"], "supporting": {"draw_boxes": {"images/img_000807.png": []}}} +{"q_idx": "detect_001544", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('airplane with white and orange livery'). For every instance, it has 1 F-16 fighter jet to its left-behind, and 1 airplane with white and orange livery to its front in view 1.\nIn view 2, locate all target instances 'F-16 fighter jet' that lie in the left-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[359.0, 367.0, 424.0, 395.0]], "image_paths": ["images/img_000282.png", "images/img_000289.png", "images/img_000290.png"], "supporting": {"draw_boxes": {"images/img_000282.png": []}}} +{"q_idx": "detect_001545", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('stylized F22 fighter jet with grey and cyan color'). For every instance, it has 1 F-22 fighter jet with gray color to its front, and 1 stylized F22 fighter jet with grey and cyan color to its left in view 1.\nIn view 2, locate all target instances 'stylized F22 fighter jet with grey and cyan color' that lie in the direct-left direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[214.0, 368.0, 326.0, 463.0]], "image_paths": ["images/img_000953.png", "images/img_000959.png"], "supporting": {"draw_boxes": {"images/img_000953.png": []}}} +{"q_idx": "detect_001546", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('green camouflage helicopter'). For every instance, it has 1 green camouflage helicopter and 1 olive green helicopter to its front in view 2.\nIn view 3, locate all target instances 'green camouflage helicopter' that lie in the right-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[270.0, 572.0, 411.0, 685.0]], "image_paths": ["images/img_001040.png", "images/img_001033.png", "images/img_001041.png"], "supporting": {"draw_boxes": {"images/img_001033.png": []}}} +{"q_idx": "detect_001547", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('lamp'). For every instance, it has 2 lamps to its front, and 1 forklift to its right-front in view 1.\nIn view 2, locate all target instances 'lamp' that lie in the direct-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[350.0, 307.0, 379.0, 352.0], [264.0, 352.0, 294.0, 400.0]], "image_paths": ["images/img_000465.png", "images/img_000473.png", "images/img_000477.png"], "supporting": {"draw_boxes": {"images/img_000465.png": []}}} +{"q_idx": "detect_001548", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('white painted wooden bench with armrests'). For every instance, it has 1 tennis ball to its right-behind in view 1.\nIn view 3, locate all target instances 'tennis ball' that lie in the left-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[226.0, 348.0, 236.0, 360.0]], "image_paths": ["images/img_000598.png", "images/img_000605.png", "images/img_000604.png"], "supporting": {"draw_boxes": {"images/img_000598.png": []}}} +{"q_idx": "detect_001549", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('airplane with red livery'). For every instance, it has 1 stylized F22 fighter jet with camoflague color to its right-front, and 1 airplane with red livery to its front in view 2.\nIn view 3, locate all target instances 'airplane with red livery' that lie in the direct-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[201.0, 435.0, 357.0, 565.0]], "image_paths": ["images/img_000274.png", "images/img_000264.png", "images/img_000269.png"], "supporting": {"draw_boxes": {"images/img_000264.png": []}}} +{"q_idx": "detect_001550", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('worker'). For every instance, it has 1 worker to its behind, and 1 white stack of sandbags to its right-front in view 1.\nIn view 3, locate all target instances 'worker' that lie in the direct-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[170.0, 535.0, 201.0, 607.0]], "image_paths": ["images/img_000453.png", "images/img_000456.png", "images/img_000459.png"], "supporting": {"draw_boxes": {"images/img_000453.png": []}}} +{"q_idx": "detect_001551", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('three balls snowman'). For every instance, it has 2 black buttons snowmans to its left in view 2.\nIn view 1, locate all target instances 'three balls snowman' that lie in the left-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[574.0, 497.0, 635.0, 610.0]], "image_paths": ["images/img_000767.png", "images/img_000760.png"], "supporting": {"draw_boxes": {"images/img_000760.png": []}}} +{"q_idx": "detect_001552", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('woman in pink clothing'). For every instance, it has 1 woman in white clothing to its right in view 2.\nIn view 1, locate all target instances 'woman in white clothing' that lie in the right-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[616.0, 392.0, 654.0, 525.0], [573.0, 423.0, 618.0, 560.0]], "image_paths": ["images/img_001054.png", "images/img_001049.png", "images/img_001058.png"], "supporting": {"draw_boxes": {"images/img_001049.png": []}}} +{"q_idx": "detect_001553", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('man wearing red jacket'). For every instance, it has 1 bus stop without enclosure to its left-front in view 2.\nIn view 3, locate all target instances 'man wearing red jacket' that lie in the left-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[460.0, 438.0, 532.0, 660.0]], "image_paths": ["images/img_000071.png", "images/img_000060.png", "images/img_000067.png"], "supporting": {"draw_boxes": {"images/img_000060.png": []}}} +{"q_idx": "detect_001554", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('stylized F22 fighter jet with grey and cyan color'). For every instance, it has 1 green and yellow striped hot air balloon to its right-front in view 1.\nIn view 2, locate all target instances 'green and yellow striped hot air balloon' that lie in the right-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[11.0, 437.0, 185.0, 700.0]], "image_paths": ["images/img_000921.png", "images/img_000927.png"], "supporting": {"draw_boxes": {"images/img_000921.png": []}}} +{"q_idx": "detect_001555", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('airplane with yellow livery'). For every instance, it has 1 airplane with white and green livery to its left in view 1.\nIn view 2, locate all target instances 'airplane with white and green livery' that lie in the direct-left direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[46.0, 453.0, 215.0, 508.0]], "image_paths": ["images/img_000251.png", "images/img_000257.png", "images/img_000261.png"], "supporting": {"draw_boxes": {"images/img_000251.png": []}}} +{"q_idx": "detect_001556", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('olive green helicopter'). For every instance, it has 1 red and black striped hot air balloon to its right-front in view 1.\nIn view 2, locate all target instances 'olive green helicopter' that lie in the direct-left direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[556.0, 338.0, 584.0, 400.0]], "image_paths": ["images/img_000986.png", "images/img_000993.png"], "supporting": {"draw_boxes": {"images/img_000986.png": []}}} +{"q_idx": "detect_001557", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('man wearing santa costume'). For every instance, it has 1 red gloves snowman to its right-front, and 1 red gloves snowman to its left-front in view 1.\nIn view 2, locate all target instances 'red gloves snowman' that lie in the left-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[603.0, 537.0, 664.0, 673.0], [566.0, 273.0, 667.0, 382.0]], "image_paths": ["images/img_000776.png", "images/img_000785.png", "images/img_000779.png"], "supporting": {"draw_boxes": {"images/img_000776.png": []}}} +{"q_idx": "detect_001558", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('green camouflage helicopter'). For every instance, it has 1 red and white helicopter to its right in view 1.\nIn view 3, locate all target instances 'red and white helicopter' that lie in the right-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[223.0, 402.0, 307.0, 490.0]], "image_paths": ["images/img_001001.png", "images/img_001006.png", "images/img_001009.png"], "supporting": {"draw_boxes": {"images/img_001001.png": []}}} +{"q_idx": "detect_001559", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('mirage fighter jet'). For every instance, it has 1 white and green striped hot air balloon to its left-behind in view 3.\nIn view 1, locate all target instances 'mirage fighter jet' that lie in the direct-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[216.0, 367.0, 347.0, 453.0]], "image_paths": ["images/img_000945.png", "images/img_000941.png", "images/img_000937.png"], "supporting": {"draw_boxes": {"images/img_000937.png": []}}} +{"q_idx": "detect_001560", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('airplane with red livery'). For every instance, it has 1 stylized F22 fighter jet with camoflague color to its right-behind, and 1 airplane with red livery to its behind in view 1.\nIn view 2, locate all target instances 'airplane with red livery' that lie in the direct-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[650.0, 480.0, 814.0, 558.0]], "image_paths": ["images/img_000264.png", "images/img_000273.png"], "supporting": {"draw_boxes": {"images/img_000264.png": []}}} +{"q_idx": "detect_001561", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('red and white helicopter'). For every instance, it has 1 red and black striped hot air balloon to its left-behind in view 2.\nIn view 1, locate all target instances 'red and black striped hot air balloon' that lie in the left-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[454.0, 152.0, 544.0, 347.0]], "image_paths": ["images/img_001009.png", "images/img_001001.png"], "supporting": {"draw_boxes": {"images/img_001001.png": []}}} +{"q_idx": "detect_001562", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('jacket snowman'). For every instance, it has 1 coupe to its left-front in view 3.\nIn view 1, locate all target instances 'coupe' that lie in the left-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[0.0, 403.0, 236.0, 560.0]], "image_paths": ["images/img_000752.png", "images/img_000751.png", "images/img_000745.png"], "supporting": {"draw_boxes": {"images/img_000745.png": []}}} +{"q_idx": "detect_001563", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('white upright tombstone with cross'). For every instance, it has 1 man in suit and tie to its right-behind in view 2.\nIn view 1, locate all target instances 'man in suit and tie' that lie in the right-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[245.0, 163.0, 314.0, 312.0]], "image_paths": ["images/img_001198.png", "images/img_001191.png"], "supporting": {"draw_boxes": {"images/img_001191.png": []}}} +{"q_idx": "detect_001564", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('person saluting'). For every instance, it has 1 blue swim ring to its left in view 1.\nIn view 3, locate all target instances 'blue swim ring' that lie in the direct-left direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[335.0, 345.0, 388.0, 395.0]], "image_paths": ["images/img_000679.png", "images/img_000688.png", "images/img_000687.png"], "supporting": {"draw_boxes": {"images/img_000679.png": []}}} +{"q_idx": "detect_001565", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('white flat tombstone with cross'). For every instance, it has 1 white upright tombstone with cross to its right-front in view 1.\nIn view 2, locate all target instances 'white upright tombstone with cross' that lie in the direct-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[221.0, 478.0, 284.0, 625.0]], "image_paths": ["images/img_001146.png", "images/img_001154.png", "images/img_001147.png"], "supporting": {"draw_boxes": {"images/img_001146.png": []}}} +{"q_idx": "detect_001566", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('woman in pink clothing'). For every instance, it has 1 food stall to its right-behind in view 1.\nIn view 2, locate all target instances 'food stall' that lie in the right-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[229.0, 252.0, 356.0, 417.0]], "image_paths": ["images/img_001094.png", "images/img_001101.png"], "supporting": {"draw_boxes": {"images/img_001094.png": []}}} +{"q_idx": "detect_001567", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('beige sand helicopter'). For every instance, it has 1 beige sand helicopter to its left, and 1 white and green striped hot air balloon to its right-behind in view 2.\nIn view 3, locate all target instances 'beige sand helicopter' that lie in the direct-left direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[190.0, 397.0, 362.0, 465.0], [49.0, 575.0, 304.0, 683.0]], "image_paths": ["images/img_000974.png", "images/img_000964.png", "images/img_000971.png"], "supporting": {"draw_boxes": {"images/img_000964.png": []}}} +{"q_idx": "detect_001568", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('desert camouflage helicopter'). For every instance, it has 1 green camouflage helicopter to its left in view 1.\nIn view 2, locate all target instances 'green camouflage helicopter' that lie in the right-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[595.0, 575.0, 740.0, 688.0]], "image_paths": ["images/img_001018.png", "images/img_001025.png"], "supporting": {"draw_boxes": {"images/img_001018.png": []}}} +{"q_idx": "detect_001569", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('man wearing green hood'). For every instance, it has 1 man wearing red jacket to its right in view 2.\nIn view 1, locate all target instances 'man wearing red jacket' that lie in the direct-right direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[270.0, 343.0, 306.0, 438.0]], "image_paths": ["images/img_000733.png", "images/img_000726.png", "images/img_000730.png"], "supporting": {"draw_boxes": {"images/img_000726.png": []}}} +{"q_idx": "detect_001570", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('black upright tombstone with cross'). For every instance, it has 1 white flat tombstone with cross to its behind in view 3.\nIn view 1, locate all target instances 'white flat tombstone with cross' that lie in the direct-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[490.0, 255.0, 616.0, 407.0]], "image_paths": ["images/img_001154.png", "images/img_001148.png", "images/img_001145.png"], "supporting": {"draw_boxes": {"images/img_001145.png": []}}} +{"q_idx": "detect_001571", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('harrier jumpjet fighter jet'). For every instance, it has 1 harrier jumpjet fighter jet to its right-front, and 1 mirage fighter jet to its right in view 1.\nIn view 2, locate all target instances 'mirage fighter jet' that lie in the right-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[216.0, 367.0, 347.0, 453.0]], "image_paths": ["images/img_000937.png", "images/img_000945.png"], "supporting": {"draw_boxes": {"images/img_000937.png": []}}} +{"q_idx": "detect_001572", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('person saluting'). For every instance, it has 1 person with black hat to its right in view 2.\nIn view 3, locate all target instances 'person with black hat' that lie in the direct-right direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[352.0, 247.0, 383.0, 393.0]], "image_paths": ["images/img_000681.png", "images/img_000679.png", "images/img_000685.png"], "supporting": {"draw_boxes": {"images/img_000679.png": []}}} +{"q_idx": "detect_001573", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('worker'). For every instance, it has 1 worker to its front, and 1 stone steps to its left-behind in view 2.\nIn view 3, locate all target instances 'stone steps' that lie in the left-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[376.0, 235.0, 555.0, 417.0]], "image_paths": ["images/img_000439.png", "images/img_000436.png", "images/img_000443.png"], "supporting": {"draw_boxes": {"images/img_000436.png": []}}} +{"q_idx": "detect_001574", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('blue painted metal bench without armrests'). For every instance, it has 1 tennis ball to its left-front, and 1 blue painted metal bench without armrests to its right in view 1.\nIn view 2, locate all target instances 'tennis ball' that lie in the left-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[145.0, 448.0, 155.0, 462.0]], "image_paths": ["images/img_000611.png", "images/img_000618.png"], "supporting": {"draw_boxes": {"images/img_000611.png": []}}} +{"q_idx": "detect_001575", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('airplane with light blue livery'). For every instance, it has 1 airplane with red livery to its behind in view 1.\nIn view 2, locate all target instances 'airplane with light blue livery' that lie in the right-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[226.0, 567.0, 420.0, 657.0], [125.0, 502.0, 306.0, 572.0]], "image_paths": ["images/img_000193.png", "images/img_000199.png", "images/img_000203.png"], "supporting": {"draw_boxes": {"images/img_000193.png": []}}} +{"q_idx": "detect_001576", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('tombstone without cross'). For every instance, it has 1 man in suit and tie to its right-behind in view 2.\nIn view 1, locate all target instances 'man in suit and tie' that lie in the right-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[884.0, 388.0, 955.0, 557.0]], "image_paths": ["images/img_001153.png", "images/img_001144.png"], "supporting": {"draw_boxes": {"images/img_001144.png": []}}} +{"q_idx": "detect_001577", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('woman in pink clothing'). For every instance, it has 1 woman in pink clothing to its front, and 1 vendor cart with red and yellow roof to its right-front in view 1.\nIn view 2, locate all target instances 'woman in pink clothing' that lie in the left-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[471.0, 488.0, 511.0, 637.0]], "image_paths": ["images/img_001098.png", "images/img_001107.png", "images/img_001102.png"], "supporting": {"draw_boxes": {"images/img_001098.png": []}}} +{"q_idx": "detect_001578", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('airplane with yellow livery'). For every instance, it has 1 airplane with white and green livery to its left in view 3.\nIn view 2, locate all target instances 'airplane with white and green livery' that lie in the direct-left direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[46.0, 453.0, 215.0, 508.0]], "image_paths": ["images/img_000259.png", "images/img_000257.png", "images/img_000251.png"], "supporting": {"draw_boxes": {"images/img_000251.png": []}}} +{"q_idx": "detect_001579", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('natural wooden bench with armrests'). For every instance, it has 2 soccer balls to its left-behind in view 2.\nIn view 1, locate all target instances 'soccer ball' that lie in the left-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[674.0, 613.0, 694.0, 640.0], [284.0, 707.0, 306.0, 735.0], [297.0, 722.0, 320.0, 752.0]], "image_paths": ["images/img_000573.png", "images/img_000564.png"], "supporting": {"draw_boxes": {"images/img_000564.png": []}}} +{"q_idx": "detect_001580", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('lamp'). For every instance, it has 1 truck with white and black cab to its left-front in view 2.\nIn view 1, locate all target instances 'truck with white and black cab' that lie in the left-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[445.0, 88.0, 559.0, 293.0]], "image_paths": ["images/img_000442.png", "images/img_000435.png", "images/img_000441.png"], "supporting": {"draw_boxes": {"images/img_000435.png": []}}} +{"q_idx": "detect_001581", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('F-16 fighter jet'). For every instance, it has 1 stylized F22 fighter jet with camoflague color to its behind in view 1.\nIn view 2, locate all target instances 'F-16 fighter jet' that lie in the right-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[318.0, 595.0, 458.0, 725.0]], "image_paths": ["images/img_000950.png", "images/img_000956.png"], "supporting": {"draw_boxes": {"images/img_000950.png": []}}} +{"q_idx": "detect_001582", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('woman in pink clothing'). For every instance, it has 1 woman wearing black dress to its front in view 2.\nIn view 3, locate all target instances 'woman in pink clothing' that lie in the direct-left direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[370.0, 387.0, 430.0, 508.0]], "image_paths": ["images/img_001067.png", "images/img_001060.png", "images/img_001069.png"], "supporting": {"draw_boxes": {"images/img_001060.png": []}}} +{"q_idx": "detect_001583", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('man wearing a dark kurta and white pants'). For every instance, it has 1 man wearing a dark kurta and white pants to its behind, and 1 colonnade with rectangular top to its left-behind in view 2.\nIn view 3, locate all target instances 'man wearing a dark kurta and white pants' that lie in the direct-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[454.0, 325.0, 475.0, 407.0]], "image_paths": ["images/img_000846.png", "images/img_000839.png", "images/img_000847.png"], "supporting": {"draw_boxes": {"images/img_000839.png": []}}} +{"q_idx": "detect_001584", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('airplane with white and orange livery'). For every instance, it has 2 airplane with white and green liveries to its behind in view 1.\nIn view 3, locate all target instances 'airplane with white and green livery' that lie in the direct-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[785.0, 453.0, 954.0, 508.0], [657.0, 413.0, 796.0, 462.0]], "image_paths": ["images/img_000219.png", "images/img_000230.png", "images/img_000225.png"], "supporting": {"draw_boxes": {"images/img_000219.png": []}}} +{"q_idx": "detect_001585", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('woman wearing black dress'). For every instance, it has 1 vendor cart with wooden roof to its left-front in view 1.\nIn view 2, locate all target instances 'vendor cart with wooden roof' that lie in the left-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[40.0, 305.0, 261.0, 553.0]], "image_paths": ["images/img_001061.png", "images/img_001067.png"], "supporting": {"draw_boxes": {"images/img_001061.png": []}}} +{"q_idx": "detect_001586", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('F-16 fighter jet'). For every instance, it has 1 stylized F22 fighter jet with grey and cyan color to its behind in view 1.\nIn view 2, locate all target instances 'stylized F22 fighter jet with grey and cyan color' that lie in the direct-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[661.0, 382.0, 777.0, 487.0]], "image_paths": ["images/img_000950.png", "images/img_000960.png"], "supporting": {"draw_boxes": {"images/img_000950.png": []}}} +{"q_idx": "detect_001587", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('airplane with yellow livery'). For every instance, it has 1 airplane with white and green livery to its behind, and 1 airplane with white and green livery to its right-behind in view 3.\nIn view 2, locate all target instances 'airplane with yellow livery' that lie in the right-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[244.0, 512.0, 409.0, 658.0]], "image_paths": ["images/img_000230.png", "images/img_000229.png", "images/img_000222.png"], "supporting": {"draw_boxes": {"images/img_000222.png": []}}} +{"q_idx": "detect_001588", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('airplane with yellow livery'). For every instance, it has 1 airplane with yellow livery to its behind, and 1 airplane with yellow livery to its front in view 2.\nIn view 3, locate all target instances 'airplane with yellow livery' that lie in the direct-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[464.0, 350.0, 564.0, 392.0]], "image_paths": ["images/img_000254.png", "images/img_000247.png", "images/img_000255.png"], "supporting": {"draw_boxes": {"images/img_000247.png": []}}} +{"q_idx": "detect_001589", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('woman wearing underwear'). For every instance, it has 1 food stall with red and white striped roof to its left-behind in view 1.\nIn view 2, locate all target instances 'food stall with red and white striped roof' that lie in the left-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[83.0, 350.0, 221.0, 557.0]], "image_paths": ["images/img_001113.png", "images/img_001119.png"], "supporting": {"draw_boxes": {"images/img_001113.png": []}}} +{"q_idx": "detect_001590", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('airplane with yellow livery'). For every instance, it has 1 airplane with red livery to its behind in view 1.\nIn view 3, locate all target instances 'airplane with red livery' that lie in the right-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[284.0, 533.0, 487.0, 637.0]], "image_paths": ["images/img_000267.png", "images/img_000273.png", "images/img_000275.png"], "supporting": {"draw_boxes": {"images/img_000267.png": []}}} +{"q_idx": "detect_001591", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('airplane with white and orange livery'). For every instance, it has 1 F-16 fighter jet to its right-front, and 1 F-16 fighter jet to its right-behind in view 3.\nIn view 1, locate all target instances 'F-16 fighter jet' that lie in the left-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[74.0, 540.0, 149.0, 588.0]], "image_paths": ["images/img_000287.png", "images/img_000285.png", "images/img_000278.png"], "supporting": {"draw_boxes": {"images/img_000278.png": []}}} +{"q_idx": "detect_001592", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('bench made of stone'). For every instance, it has 1 soccer ball to its left-behind in view 3.\nIn view 2, locate all target instances 'soccer ball' that lie in the left-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[588.0, 303.0, 601.0, 322.0]], "image_paths": ["images/img_000620.png", "images/img_000623.png", "images/img_000615.png"], "supporting": {"draw_boxes": {"images/img_000615.png": []}}} +{"q_idx": "detect_001593", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('woman in white clothing'). For every instance, it has 1 woman wearing scarf accessory to its front in view 1.\nIn view 2, locate all target instances 'woman wearing scarf accessory' that lie in the left-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[384.0, 458.0, 412.0, 597.0]], "image_paths": ["images/img_001096.png", "images/img_001104.png"], "supporting": {"draw_boxes": {"images/img_001096.png": []}}} +{"q_idx": "detect_001594", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 1, find all reference instances ('worker'). For every instance, it has 1 worker to its behind, and 1 white stack of sandbags to its right-front in view 1.\nIn view 2, locate all target instances 'worker' that lie in the direct-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[224.0, 368.0, 251.0, 430.0]], "image_paths": ["images/img_000453.png", "images/img_000460.png"], "supporting": {"draw_boxes": {"images/img_000453.png": []}}} +{"q_idx": "detect_001595", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('F-22 fighter jet with gray color'). For every instance, it has 1 stylized F22 fighter jet with camoflague color to its front in view 2.\nIn view 1, locate all target instances 'F-22 fighter jet with gray color' that lie in the right-behind direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[224.0, 383.0, 334.0, 460.0]], "image_paths": ["images/img_000911.png", "images/img_000902.png", "images/img_000913.png"], "supporting": {"draw_boxes": {"images/img_000902.png": []}}} +{"q_idx": "detect_001596", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('woman in gray clothing'). For every instance, it has 1 woman wearing underwear to its left in view 3.\nIn view 2, locate all target instances 'woman wearing underwear' that lie in the direct-right direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[381.0, 357.0, 412.0, 503.0]], "image_paths": ["images/img_001117.png", "images/img_001122.png", "images/img_001114.png"], "supporting": {"draw_boxes": {"images/img_001114.png": []}}} +{"q_idx": "detect_001597", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('beige tombstone'). For every instance, it has 1 man taking off white shorts to its right-front in view 2.\nIn view 1, locate all target instances 'man taking off white shorts' that lie in the right-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[246.0, 190.0, 295.0, 330.0]], "image_paths": ["images/img_001164.png", "images/img_001158.png"], "supporting": {"draw_boxes": {"images/img_001158.png": []}}} +{"q_idx": "detect_001598", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('man in a plaid shirt with his arms crossed'). For every instance, it has 1 man holding box to its front in view 2.\nIn view 3, locate all target instances 'man holding box' that lie in the direct-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 3.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[681.0, 408.0, 714.0, 502.0]], "image_paths": ["images/img_000878.png", "images/img_000874.png", "images/img_000883.png"], "supporting": {"draw_boxes": {"images/img_000874.png": []}}} +{"q_idx": "detect_001599", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 2, find all reference instances ('man wearing santa costume'). For every instance, it has 1 man wearing santa costume to its front, and 1 three balls snowman to its right-front in view 2.\nIn view 1, locate all target instances 'man wearing santa costume' that lie in the direct-front direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 1.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[398.0, 240.0, 432.0, 323.0]], "image_paths": ["images/img_000801.png", "images/img_000794.png", "images/img_000800.png"], "supporting": {"draw_boxes": {"images/img_000794.png": []}}} +{"q_idx": "detect_001600", "q_type": "detect", "question": "All views are captured from the same static scene arrangement, but from different positions and angles.\nIn view 3, find all reference instances ('woman in white clothing'). For every instance, it has 1 woman wearing black dress to its left, and 1 woman in white clothing to its right in view 3.\nIn view 2, locate all target instances 'woman wearing black dress' that lie in the direct-left direction relative to these reference instances.\nThen return the bounding boxes of those target instances in view 2.\n", "options": {}, "question_prompt": "Return a single flat list of bounding boxes, one bounding box per target instance. Each bounding box is in normalized xyxy format [x1, y1, x2, y2]. If no such instances exist, return an empty list.", "answer": [[395.0, 323.0, 434.0, 462.0]], "image_paths": ["images/img_001069.png", "images/img_001067.png", "images/img_001061.png"], "supporting": {"draw_boxes": {"images/img_001061.png": []}}} diff --git a/images/img_000001.png b/images/img_000001.png new file mode 100644 index 0000000000000000000000000000000000000000..1ec254fcee117f59ecaba3b6b59bb6097adc21e4 --- /dev/null +++ b/images/img_000001.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f62aa0a84e3e92356d853a24b8fb4f0d8a4bdcf09ef6b873d1f5ad8f1c9cdd85 +size 660047 diff --git a/images/img_000002.png b/images/img_000002.png new file mode 100644 index 0000000000000000000000000000000000000000..d551485d9ea82014a1cfed9b82b6eeafc5445d67 --- /dev/null +++ b/images/img_000002.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:66233186f1ac11c275d4447f485f03795b84b55fa594c5c47d1ae0100cc428e8 +size 641205 diff --git a/images/img_000003.png b/images/img_000003.png new file mode 100644 index 0000000000000000000000000000000000000000..daa8b4e83b92daa4c5d56158f26839f4577c9cfc --- /dev/null +++ b/images/img_000003.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1976b150db1b3bc26de869974a59168f5c528b9f7714e8c6f8180914fbb86b5d +size 610232 diff --git a/images/img_000004.png b/images/img_000004.png new file mode 100644 index 0000000000000000000000000000000000000000..d81a77b0140fc231ec9ff5aefc6b2d401ca58a98 --- /dev/null +++ b/images/img_000004.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2b9fed6bc2c20fd77589d4ab8b04d8d33c1305cc6652ad7dcbbca03319b3fdbf +size 683340 diff --git a/images/img_000005.png b/images/img_000005.png new file mode 100644 index 0000000000000000000000000000000000000000..53d237f568293de84e852c142550b137f0ad095d --- /dev/null +++ b/images/img_000005.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:aec2e1f5403a03255dcbc269d6285e76e72ae3b633a5393934c84f1493024e59 +size 661525 diff --git a/images/img_000006.png b/images/img_000006.png new file mode 100644 index 0000000000000000000000000000000000000000..f62adf7da882f88b956e5166ff501c09c1d077e1 --- /dev/null +++ b/images/img_000006.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:89d56c5b83ac5187c41cef9d97b2e8e7d579a8a05f9244d05bc30aa9666b3838 +size 658938 diff --git a/images/img_000007.png b/images/img_000007.png new file mode 100644 index 0000000000000000000000000000000000000000..1ebe551dc65514ae5f84bffaf057e4f217023012 --- /dev/null +++ b/images/img_000007.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0d2c5f1a5425268c0e9e99cc52aa0232ef9a53a8ff3b17a35cf1c84b60a6d1bb +size 680313 diff --git a/images/img_000008.png b/images/img_000008.png new file mode 100644 index 0000000000000000000000000000000000000000..e06efe4e63d3e5462e2c6555db6c648a5ef359cd --- /dev/null +++ b/images/img_000008.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:abda16a5a01fd439c0e06ffc60316ab1b4276f85da3fb0f7b4cb5a7cbc8356d7 +size 746344 diff --git a/images/img_000009.png b/images/img_000009.png new file mode 100644 index 0000000000000000000000000000000000000000..d8a68346ec8e0d55514e7d92d327b57443c67f73 --- /dev/null +++ b/images/img_000009.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:57f9d2fcd26b066dde39492d8cad5dc4aaf3dffd867053815479126cd892b91f +size 668693 diff --git a/images/img_000010.png b/images/img_000010.png new file mode 100644 index 0000000000000000000000000000000000000000..fdda94b44ea4fbb5ade67d02de8c83179fcef158 --- /dev/null +++ b/images/img_000010.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7d859e2e737209fb8237cc0b4e8123fd67188ddd8e15a4b9b55b326e4ae41456 +size 777041 diff --git a/images/img_000011.png b/images/img_000011.png new file mode 100644 index 0000000000000000000000000000000000000000..a66ca04f34e022d9d19c097e13c795f8348e7242 --- /dev/null +++ b/images/img_000011.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1efd1522c239b32c7f83706136ca0bd15ed499bc67a1a7be2ef5451405a341a2 +size 786190 diff --git a/images/img_000012.png b/images/img_000012.png new file mode 100644 index 0000000000000000000000000000000000000000..d8cd5c944f0a32a9d90fb7883b37881e4c03fe53 --- /dev/null +++ b/images/img_000012.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ffc0cc6104aeb30e2b85bb6daf1962730b761558a4d411f83696e44b0e110e50 +size 783575 diff --git a/images/img_000013.png b/images/img_000013.png new file mode 100644 index 0000000000000000000000000000000000000000..2bd8e8a6f0b7fe28d26256c184f8abc6f9680e06 --- /dev/null +++ b/images/img_000013.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9b8900b3517bf1b6d1393a27c7603e8a64d644f5f4a3fcff268aa480ebdfe1ab +size 756318 diff --git a/images/img_000014.png b/images/img_000014.png new file mode 100644 index 0000000000000000000000000000000000000000..f8c774b8e78f71de19295ebf69514a67ae50b6a7 --- /dev/null +++ b/images/img_000014.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:efa915c21d56eb89947b5bd70ba92251eba0f7e00459221b5639e6f99845ee79 +size 779475 diff --git a/images/img_000015.png b/images/img_000015.png new file mode 100644 index 0000000000000000000000000000000000000000..4cf067939bd96edc6a48c773d08c7ff5372b31b8 --- /dev/null +++ b/images/img_000015.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4103b073ef2a65ff1bfd16be610763988402617e34f8f802de5e7ed6510cc706 +size 437702 diff --git a/images/img_000016.png b/images/img_000016.png new file mode 100644 index 0000000000000000000000000000000000000000..350e632b0e43cf0a74013697b78d8f382c4c528c --- /dev/null +++ b/images/img_000016.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1075ccf537fe4a802c7dbdd20184b186e8783c8c2594f220991289e32b240a00 +size 638490 diff --git a/images/img_000017.png b/images/img_000017.png new file mode 100644 index 0000000000000000000000000000000000000000..029b2eee3a75a83f009c3be36fd053c93d397692 --- /dev/null +++ b/images/img_000017.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3390897fa542d72b80909594db7339ec7575396de1279a2dc628741532d463c3 +size 625527 diff --git a/images/img_000018.png b/images/img_000018.png new file mode 100644 index 0000000000000000000000000000000000000000..2187e7e948f8bcfbeca55aeeebcf9016b5c21e69 --- /dev/null +++ b/images/img_000018.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:94630ccaf23edb1cf6a88349ea7d6bb24a03d5ce90c86db322f752fdd6612c88 +size 621991 diff --git a/images/img_000019.png b/images/img_000019.png new file mode 100644 index 0000000000000000000000000000000000000000..8fe2410f0be4a9cb676f7e48868b6fa30a9f9112 --- /dev/null +++ b/images/img_000019.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b7c0ea4bea8e1dcb32940e2bd28db8f35935a4d7223fc7abf934e552a945b104 +size 644595 diff --git a/images/img_000020.png b/images/img_000020.png new file mode 100644 index 0000000000000000000000000000000000000000..f4564f9cd2b81b82268c822c74f3c66c49bba2a1 --- /dev/null +++ b/images/img_000020.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6052978f546545c9e6cfddabef06811d4994f2e89f056ee052f0180bccc7450a +size 598633 diff --git a/images/img_000021.png b/images/img_000021.png new file mode 100644 index 0000000000000000000000000000000000000000..817ad1a474170416774351311d6d654acb552c36 --- /dev/null +++ b/images/img_000021.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0cbf4ee550117e62abea2d97f86c1c5e74558685dcd4164a2f78d906f3854974 +size 656814 diff --git a/images/img_000022.png b/images/img_000022.png new file mode 100644 index 0000000000000000000000000000000000000000..79890ad83773b2fbc8a65bf4d3cbe5250d6c79a7 --- /dev/null +++ b/images/img_000022.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6b7eaef99390ce33fd58ca112e3dc022517f286d71800489dfda871a3ed69b06 +size 706293 diff --git a/images/img_000023.png b/images/img_000023.png new file mode 100644 index 0000000000000000000000000000000000000000..b8afba75342e57621081097e0688292e0bb76e8a --- /dev/null +++ b/images/img_000023.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:10e88af91f16b3835e1e79fc0f769bb9780a38961bca29a79a032e17ae9a65e4 +size 686869 diff --git a/images/img_000024.png b/images/img_000024.png new file mode 100644 index 0000000000000000000000000000000000000000..0fd6430bb3eddc8b9ce8017ecafe02497088c7d9 --- /dev/null +++ b/images/img_000024.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:818148d4f2b7fdc6b0712d011d7aaa72e206cfe0e1c3da4ee1ea84185a2ec2f2 +size 676055 diff --git a/images/img_000025.png b/images/img_000025.png new file mode 100644 index 0000000000000000000000000000000000000000..7f06b23f45372c97e5dd07afd9459a3885ebc27a --- /dev/null +++ b/images/img_000025.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6b42ccc171d8ea8f1ab184e9309f0d6f41f0e2d0a54f5182f95f1bd55d188c94 +size 600169 diff --git a/images/img_000026.png b/images/img_000026.png new file mode 100644 index 0000000000000000000000000000000000000000..ad71784ffc1b3e8ea36fc83eaddda2d362b1a6ce --- /dev/null +++ b/images/img_000026.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c82020631d822c20d40aa8ef5cc1f3837ed1520bd76f59d79fccb85829c496e3 +size 688822 diff --git a/images/img_000027.png b/images/img_000027.png new file mode 100644 index 0000000000000000000000000000000000000000..159af53cede86ec06e9e5643309908d02b2a99d5 --- /dev/null +++ b/images/img_000027.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:30659839a9a089cb3e573d9b6b3083e98821d642128a394ea5507d205a04ad29 +size 687130 diff --git a/images/img_000028.png b/images/img_000028.png new file mode 100644 index 0000000000000000000000000000000000000000..e2580fc60aa1b6f6043119badbf1a879a4b97fd9 --- /dev/null +++ b/images/img_000028.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8ecb03829e97dda77d8a442b06f92e1b1d43343bba1fabf11a227e6b053606ec +size 690047 diff --git a/images/img_000029.png b/images/img_000029.png new file mode 100644 index 0000000000000000000000000000000000000000..d12eae6fdf3a0818df82d616955023dbeb634694 --- /dev/null +++ b/images/img_000029.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:daeee1b9d603eee811c5591d49c030035633e260b414f2e1198920f1afe5fa87 +size 683064 diff --git a/images/img_000030.png b/images/img_000030.png new file mode 100644 index 0000000000000000000000000000000000000000..7b16e96c0796a53d804e72d0fa141a9995b46986 --- /dev/null +++ b/images/img_000030.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bdce4c491097c32b4dea40bf24dd80cc037e0388a5bac73e5ee3239c7b77c2d4 +size 600071 diff --git a/images/img_000031.png b/images/img_000031.png new file mode 100644 index 0000000000000000000000000000000000000000..ba962979f34d141eeb96a0ac1fbd6e75d58ff28c --- /dev/null +++ b/images/img_000031.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:be508b69168d08d8fd6ff4c24deb46e9fdec44120490f1711e6f5f6250b1e9f0 +size 614406 diff --git a/images/img_000032.png b/images/img_000032.png new file mode 100644 index 0000000000000000000000000000000000000000..f9ee5efbf5af62c63e1d59f71bf9af03306ddb25 --- /dev/null +++ b/images/img_000032.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1f39248e53cf322e4b858368f375bf13a65f0ae6d339a34f743bc3d615713b11 +size 614913 diff --git a/images/img_000033.png b/images/img_000033.png new file mode 100644 index 0000000000000000000000000000000000000000..b442a53935ad1ebcdff756b29184b43f50003f30 --- /dev/null +++ b/images/img_000033.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:96d4ee457a48cca5728e52d4e3d4b43272c94af484d2c8aa7fb6636c23087f27 +size 580891 diff --git a/images/img_000034.png b/images/img_000034.png new file mode 100644 index 0000000000000000000000000000000000000000..f295ecfcc63263379c629caafefbbd850e82ff23 --- /dev/null +++ b/images/img_000034.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:406d9c2f0348240b8c3980b1668570721d93a7d7e31662f56f74a4f179f7b128 +size 543649 diff --git a/images/img_000035.png b/images/img_000035.png new file mode 100644 index 0000000000000000000000000000000000000000..5c10ececf14abc5eadebe63b3b605182332e0244 --- /dev/null +++ b/images/img_000035.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1227e6bd3c0a0fffeb62ce285fa427a331c99aea18377bcfa5745925c9b7efa3 +size 554344 diff --git a/images/img_000036.png b/images/img_000036.png new file mode 100644 index 0000000000000000000000000000000000000000..bc6499c56cd55bf78400a9cf8178bdef9f2315e1 --- /dev/null +++ b/images/img_000036.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:19b337e102d9342c859aeb184799c2c64e58145f19ba895158b77595c8dbc218 +size 615889 diff --git a/images/img_000037.png b/images/img_000037.png new file mode 100644 index 0000000000000000000000000000000000000000..d2b8a4d8328587214037324e7d2f773b2d409ac4 --- /dev/null +++ b/images/img_000037.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0fb0d44a5f26834ad2cc5e6b534d9f9fb691227c9d07d1f70ec0aa83ec043bf2 +size 664058 diff --git a/images/img_000038.png b/images/img_000038.png new file mode 100644 index 0000000000000000000000000000000000000000..b8c366505e81258cd51f2d9b068a94258df14341 --- /dev/null +++ b/images/img_000038.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5c9a52e16f41825f022e62796017d64d4458452f6411218b5401fab000092532 +size 627874 diff --git a/images/img_000039.png b/images/img_000039.png new file mode 100644 index 0000000000000000000000000000000000000000..7e99438d0870c2c5a37e06ced4561e385ee6d1e1 --- /dev/null +++ b/images/img_000039.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:23c4c61f14ef8d0089c48ad10cca41b3e817b7666e7145a1b5a4a5f62ae23527 +size 644720 diff --git a/images/img_000040.png b/images/img_000040.png new file mode 100644 index 0000000000000000000000000000000000000000..619a160c6a3d9b8f03714c12085f288144b44269 --- /dev/null +++ b/images/img_000040.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ca542da56ae58dcd5e7e3cac27f44b2b9acc0aa7036153483a5a10ad05054d2a +size 608163 diff --git a/images/img_000041.png b/images/img_000041.png new file mode 100644 index 0000000000000000000000000000000000000000..a2e321a751b1bc25a6bea5090b57a60da5981985 --- /dev/null +++ b/images/img_000041.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:458b08b1022a5da3cbe7438c182717ae3d97b6ba5bb238119eef5fc2debb55ab +size 622507 diff --git a/images/img_000042.png b/images/img_000042.png new file mode 100644 index 0000000000000000000000000000000000000000..6b3bc445e4bd46ec0cb7423dc50e714b874768d5 --- /dev/null +++ b/images/img_000042.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7c7cdb31dffb734e3bc86659eda02759c54b158d792b1e6f3038c97517ea538c +size 628185 diff --git a/images/img_000043.png b/images/img_000043.png new file mode 100644 index 0000000000000000000000000000000000000000..541d97b44b0137a0000ed315eacf570b5a3f33fa --- /dev/null +++ b/images/img_000043.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:07a71e6cdce3f361da948e739ad255674d3d1dc36ad761af2a67ca8bcb7a95c9 +size 656213 diff --git a/images/img_000044.png b/images/img_000044.png new file mode 100644 index 0000000000000000000000000000000000000000..0cb056c27af3912ec76fbaf3bd28d90f7495e1e0 --- /dev/null +++ b/images/img_000044.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2246719bfaaaa218f6ae949b6a1074bf420bb9364c7b1812209d409a15136e78 +size 624400 diff --git a/images/img_000045.png b/images/img_000045.png new file mode 100644 index 0000000000000000000000000000000000000000..e039cdf59b224536db93f5d631e721ef18b5918d --- /dev/null +++ b/images/img_000045.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:29991998ba23ea73548b32b1327a7f3e4edd60d9e39798018fbaa20aa19c9ed1 +size 519177 diff --git a/images/img_000046.png b/images/img_000046.png new file mode 100644 index 0000000000000000000000000000000000000000..bf725504cf9daedc6ad44b9a8ca5c3231a97b4ab --- /dev/null +++ b/images/img_000046.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ec774af1edddff5d051a69b509371ef568998027a7e8d1ab4931a311d053ebf0 +size 596373 diff --git a/images/img_000047.png b/images/img_000047.png new file mode 100644 index 0000000000000000000000000000000000000000..f620374a5d0dacdd210935f78a37a8f9b32a8e2b --- /dev/null +++ b/images/img_000047.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dc764deab6f34d7928c7b158fd861f267f0502b477550c4e5ec6a4981590a4be +size 618399 diff --git a/images/img_000048.png b/images/img_000048.png new file mode 100644 index 0000000000000000000000000000000000000000..834611638089bfc98da937238983c8074aff8b1c --- /dev/null +++ b/images/img_000048.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ebe619dbafb1c35d01e8970ce161736e636284378fe2479470a6fe2c1993804e +size 588665 diff --git a/images/img_000049.png b/images/img_000049.png new file mode 100644 index 0000000000000000000000000000000000000000..26b43877829471512206d29227f7faf3ca590ed8 --- /dev/null +++ b/images/img_000049.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4d57aa8c3b30d95c29109d3bc8a47e20847177463cb26f21125762234f46b8bd +size 607938 diff --git a/images/img_000050.png b/images/img_000050.png new file mode 100644 index 0000000000000000000000000000000000000000..85b2bfb537d0ee8bc4c293935444926d86fa1304 --- /dev/null +++ b/images/img_000050.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:03d34999bca784b19962fb2c6d15a6e1f866685339a28fd266361dc20c1213a4 +size 624397 diff --git a/images/img_000051.png b/images/img_000051.png new file mode 100644 index 0000000000000000000000000000000000000000..95b9a2a3fb9ad526bb167bc11c0cb237571a695c --- /dev/null +++ b/images/img_000051.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9ebc7dc56bd330f688dfe08a7ca46b53c0d7f1b51c6ea9a2a133ae09290c2ad8 +size 640544 diff --git a/images/img_000052.png b/images/img_000052.png new file mode 100644 index 0000000000000000000000000000000000000000..f351f6d0e511d44a8d5d71fc08caf23c5aa91747 --- /dev/null +++ b/images/img_000052.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7e0998bca0e02c88f913373dd3efc177a437cd8f13b57a48e5976acfa8373474 +size 597585 diff --git a/images/img_000053.png b/images/img_000053.png new file mode 100644 index 0000000000000000000000000000000000000000..16e2322a9a8ca5724a12a9c3043b3f0418599950 --- /dev/null +++ b/images/img_000053.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f4f3482c5a88ef86e340b5a9c2b9e951a85da072bfb25fcc9150caec5f1d535d +size 619841 diff --git a/images/img_000054.png b/images/img_000054.png new file mode 100644 index 0000000000000000000000000000000000000000..ce1e118bcb56ab53d716961fbd8b15a4febdc858 --- /dev/null +++ b/images/img_000054.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9a9d402d178df9cf6673fb632cf279a9175ba36664fba70d59531d1043edb031 +size 575337 diff --git a/images/img_000055.png b/images/img_000055.png new file mode 100644 index 0000000000000000000000000000000000000000..abb28ac1c2a34ae6ee315999b550425607416ecf --- /dev/null +++ b/images/img_000055.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b6bd71046e839efe3614d3bdc4bd4bec91e5e743ffaa67ab82e405a3716dda38 +size 606271 diff --git a/images/img_000056.png b/images/img_000056.png new file mode 100644 index 0000000000000000000000000000000000000000..55cf75c0e826819bffecbac2af1ddf5e8acaf8ea --- /dev/null +++ b/images/img_000056.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e2c68299c99ff2143c15370db04efc280df4baa7d7e1963c643e6827258d1c37 +size 609122 diff --git a/images/img_000057.png b/images/img_000057.png new file mode 100644 index 0000000000000000000000000000000000000000..5befbad9bc82650baf5ff4346ec6d8f1379a0ced --- /dev/null +++ b/images/img_000057.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8a04241e63426e268b8a757f1863047f2087554dda9a15d650a4d676125719ad +size 634308 diff --git a/images/img_000058.png b/images/img_000058.png new file mode 100644 index 0000000000000000000000000000000000000000..d46cf98cad44ac88ec6f73d017241f310b42e4d9 --- /dev/null +++ b/images/img_000058.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ac4b9ab5b3ff5f5c65e01d810d3a5da80632858cf718c3183399f575499cd219 +size 641587 diff --git a/images/img_000059.png b/images/img_000059.png new file mode 100644 index 0000000000000000000000000000000000000000..c488e294168b714f8c29642957afbfae70e8bae5 --- /dev/null +++ b/images/img_000059.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d2970535c8d65d3c927496196600959a047ab366886dbff4e1bdd830f2254d4c +size 631115 diff --git a/images/img_000060.png b/images/img_000060.png new file mode 100644 index 0000000000000000000000000000000000000000..6c31b341821ffcbeec71ae890623b3db15c698e2 --- /dev/null +++ b/images/img_000060.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d7521f07a1034a109693ab5fe517b3e079e281d223551fefae4c1ed6cab9a99f +size 572952 diff --git a/images/img_000061.png b/images/img_000061.png new file mode 100644 index 0000000000000000000000000000000000000000..af6aad4304b64007a2e1d5a16e3739ec1a6db0f0 --- /dev/null +++ b/images/img_000061.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0f85efc79ae0851d4b0c63f81254238e15c41285fa44366e21e498f7fe78c8af +size 592447 diff --git a/images/img_000062.png b/images/img_000062.png new file mode 100644 index 0000000000000000000000000000000000000000..6f9ed6c0a7a8ce944e27500a0247269cb77fce39 --- /dev/null +++ b/images/img_000062.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6626cfce89f6e0cf22eac08a2936c85062ffe20fbf278f072406ebddfd6e6c9e +size 592559 diff --git a/images/img_000063.png b/images/img_000063.png new file mode 100644 index 0000000000000000000000000000000000000000..8749a00356dffdf4477ca8847c1d0a4d94e36b91 --- /dev/null +++ b/images/img_000063.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:24c6a93a4358b69b7ae9940a47faa00fdc1228752958db5e4f92847c25e0677c +size 536568 diff --git a/images/img_000064.png b/images/img_000064.png new file mode 100644 index 0000000000000000000000000000000000000000..2997bbc81882b52546fc6e04b0a0bcb67e035a68 --- /dev/null +++ b/images/img_000064.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e9fbd1355845f062eab056dd5fae119af89f649e118b86d3e1209bbc5cacdce1 +size 575971 diff --git a/images/img_000065.png b/images/img_000065.png new file mode 100644 index 0000000000000000000000000000000000000000..1b97f353b5cdd37e99ac18bf49fa19628f046433 --- /dev/null +++ b/images/img_000065.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cbdabc9fc23b151ef261fd7e6b6c4b2f70dec7bb8a6332c1916fe5a212494998 +size 653504 diff --git a/images/img_000066.png b/images/img_000066.png new file mode 100644 index 0000000000000000000000000000000000000000..261f5dfd0a83a93c7cf07fcc388f3be5af1cf5fb --- /dev/null +++ b/images/img_000066.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:42cc3657c239dd6e63162e78692671fdc86bd4ce13d0092a3041ae714e8d55c6 +size 689847 diff --git a/images/img_000067.png b/images/img_000067.png new file mode 100644 index 0000000000000000000000000000000000000000..5c11b815c28f96223368043f83cd8bf2c81fa99d --- /dev/null +++ b/images/img_000067.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1c286605cd358d91c2cae7464c1526549138da779a1678f176dcbb24600442f2 +size 662656 diff --git a/images/img_000068.png b/images/img_000068.png new file mode 100644 index 0000000000000000000000000000000000000000..1a4b677fca7cf7274781e840dc0e9ad8ad34d741 --- /dev/null +++ b/images/img_000068.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0d91e81dcc650f40106a5e2ee206c85120beb728621f529f8562fa91dc87b2b8 +size 670020 diff --git a/images/img_000069.png b/images/img_000069.png new file mode 100644 index 0000000000000000000000000000000000000000..98ef50447f76f13868761c96c8d11f3c219032dd --- /dev/null +++ b/images/img_000069.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6cc23319aaf9d6df5099319a3726b1153dba95453d91a247df566d1077c174a1 +size 573369 diff --git a/images/img_000070.png b/images/img_000070.png new file mode 100644 index 0000000000000000000000000000000000000000..36a6d589d0e54d120b01375578a0450199132c45 --- /dev/null +++ b/images/img_000070.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6a6ab69bf337942124f193873a8f2f7508857f949f0efced8933b2ab36bf9f9c +size 654204 diff --git a/images/img_000071.png b/images/img_000071.png new file mode 100644 index 0000000000000000000000000000000000000000..eac731ba5aca350d0340661d181afb4a1e59ddad --- /dev/null +++ b/images/img_000071.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7068418c264692ff5d99a0ab77e342dd1a7f486ee122312aa49d460dfea82938 +size 667108 diff --git a/images/img_000072.png b/images/img_000072.png new file mode 100644 index 0000000000000000000000000000000000000000..81d1d1a066294450812896b922150a80540152e1 --- /dev/null +++ b/images/img_000072.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d730075099d744e6ad2e5ddf02362732ce16e2dca1e7521fee78ab0108262bfb +size 703506 diff --git a/images/img_000073.png b/images/img_000073.png new file mode 100644 index 0000000000000000000000000000000000000000..72ea5488494859b220603ec3b97132ed4fc40837 --- /dev/null +++ b/images/img_000073.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:47ec120e800d38c769fdbddeed882fb2da957a00cc93b7b24c99bd4906c6a055 +size 660081 diff --git a/images/img_000074.png b/images/img_000074.png new file mode 100644 index 0000000000000000000000000000000000000000..a1899b3904babba39cf96c98679cd96d004f9516 --- /dev/null +++ b/images/img_000074.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e993e31fbe30c159338f7e849682746b37fab47002ed00ef2e77b7539aae3310 +size 590309 diff --git a/images/img_000075.png b/images/img_000075.png new file mode 100644 index 0000000000000000000000000000000000000000..707cfc6e091deab076f8e64441f46aaecd2dd41b --- /dev/null +++ b/images/img_000075.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:893b7194ded8dfd27d7c301b0b46754a96210f92a5ce817d5e03f451b332c6d2 +size 530611 diff --git a/images/img_000076.png b/images/img_000076.png new file mode 100644 index 0000000000000000000000000000000000000000..814054eac359b0432c8fd4cad3631a35ee2cc76a --- /dev/null +++ b/images/img_000076.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:22f08527fa6d56a7d2cbe7ab29391f2a0dfa5525ae9b1f95deaa6d1d66040a68 +size 551963 diff --git a/images/img_000077.png b/images/img_000077.png new file mode 100644 index 0000000000000000000000000000000000000000..93b3c2469ccec5fe3c49fa494146614f71f605be --- /dev/null +++ b/images/img_000077.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7801882dba68c8cbecfd1325c4eacd28af460832ee530d05a7e74366b2ebc388 +size 541124 diff --git a/images/img_000078.png b/images/img_000078.png new file mode 100644 index 0000000000000000000000000000000000000000..b118c1393b08e1f365ca8ccccaf6e3b10e3069ac --- /dev/null +++ b/images/img_000078.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:47a968a0cdde96b7be9a4a763b96975de260ec0a169566b3e7d270dd71ac8f06 +size 728829 diff --git a/images/img_000079.png b/images/img_000079.png new file mode 100644 index 0000000000000000000000000000000000000000..984a8321e193fde44b5bbb7ca9dff32a340f7187 --- /dev/null +++ b/images/img_000079.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a83499f911beff4e2bfc2fe2e05491bbd8dae9539c295ed1ada280e5ca7b9174 +size 776328 diff --git a/images/img_000080.png b/images/img_000080.png new file mode 100644 index 0000000000000000000000000000000000000000..6e0898e28ee2673a7899e210117cea9750bbfa0a --- /dev/null +++ b/images/img_000080.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b16b82c07fbef8f2072cf8c434493fa42768d5414dbebddc5f3d7cc50cfd341a +size 726811 diff --git a/images/img_000081.png b/images/img_000081.png new file mode 100644 index 0000000000000000000000000000000000000000..7c437b220461f1e3bb2af74fc5fa589345130f2f --- /dev/null +++ b/images/img_000081.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:edd4449aa7f80472b7b0ea819e561fc3aed2fc1e4187f5fe268cb364a2ac43c5 +size 790379 diff --git a/images/img_000082.png b/images/img_000082.png new file mode 100644 index 0000000000000000000000000000000000000000..213bba855e8deace60c5abd67f99a6da34a90105 --- /dev/null +++ b/images/img_000082.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:44183d3d141e057219d5883a87f83ce56afe1d919de7331f998909ba5508b77e +size 764533 diff --git a/images/img_000083.png b/images/img_000083.png new file mode 100644 index 0000000000000000000000000000000000000000..77f68f525fc1ea0c339bc260f1daaa256c3ef8a6 --- /dev/null +++ b/images/img_000083.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b4fa6511a3f8de6c6866c4643bf369c3039e530f5a22521a17ded4314840fac7 +size 809045 diff --git a/images/img_000084.png b/images/img_000084.png new file mode 100644 index 0000000000000000000000000000000000000000..95a1377ddd24db21e530e89067c9e999315ea47d --- /dev/null +++ b/images/img_000084.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d077689c9b3973464d79e660cc3d4764d522b802a226237fa270026a338374b3 +size 784482 diff --git a/images/img_000085.png b/images/img_000085.png new file mode 100644 index 0000000000000000000000000000000000000000..c3604cc9f4969571e6e4de61d9213b0797393ef1 --- /dev/null +++ b/images/img_000085.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b07652d7df692e76e2e4c2eabe95afeffe28472cba1e57ad240e279a0300daf5 +size 828031 diff --git a/images/img_000086.png b/images/img_000086.png new file mode 100644 index 0000000000000000000000000000000000000000..b07486ad011e3893b1b07e3ae7f9379d9e6ef022 --- /dev/null +++ b/images/img_000086.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:03c7d486e052ab10fd11c250326ab8b70f73a4ae6560e4138a8fb8a7118592fc +size 699920 diff --git a/images/img_000087.png b/images/img_000087.png new file mode 100644 index 0000000000000000000000000000000000000000..7cda3c1d79c3bd1869eedcf153792d44a462a013 --- /dev/null +++ b/images/img_000087.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8b0acf26ec0f61f0c9eca21e9fadcb5502a9d148712958ce38635ce42ce6eda0 +size 565966 diff --git a/images/img_000088.png b/images/img_000088.png new file mode 100644 index 0000000000000000000000000000000000000000..4a295099ea4d1c1d8675f63a55e04df336d6dd99 --- /dev/null +++ b/images/img_000088.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3549a0b515e27f0dee1dc1116afc3074b9c02ba1b3822701c92e063dd82b4ccb +size 566006 diff --git a/images/img_000089.png b/images/img_000089.png new file mode 100644 index 0000000000000000000000000000000000000000..1a6da7f36f7755cd6a7149616e7c08953886995c --- /dev/null +++ b/images/img_000089.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0e34f20d0bfff218bdd11f8691dbe5ab76363f5d3a75cde59b46e7ff6c9c1a75 +size 568664 diff --git a/images/img_000090.png b/images/img_000090.png new file mode 100644 index 0000000000000000000000000000000000000000..fc9869c06960738f6da9a64e0ad53c1931cc2dbe --- /dev/null +++ b/images/img_000090.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c89a2ea1e2f7e8f54511d2095f606c9c98ac6b80da45cdec8446183829790ffc +size 699807 diff --git a/images/img_000091.png b/images/img_000091.png new file mode 100644 index 0000000000000000000000000000000000000000..4f69b874dfe027fd28a36a00d40bc63a1e925ff2 --- /dev/null +++ b/images/img_000091.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:26bc4816447d03d05cbd7dccdb13e3c0a5f6fce4d0ca67313c5df9854177d049 +size 771161 diff --git a/images/img_000092.png b/images/img_000092.png new file mode 100644 index 0000000000000000000000000000000000000000..24e81c681fffc86ee2354825028838c7cb3303bb --- /dev/null +++ b/images/img_000092.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3da6be70adff909eb78998a1fc8dbb36c607aaf75258403aba829c4d2281de63 +size 753547 diff --git a/images/img_000093.png b/images/img_000093.png new file mode 100644 index 0000000000000000000000000000000000000000..eeeba1c745f8af9e2aa2e8876f7aebca2949a593 --- /dev/null +++ b/images/img_000093.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:12a103842808846ccdae9669c389c1e731b7651c2f2ec874623febc3569eb1de +size 811903 diff --git a/images/img_000094.png b/images/img_000094.png new file mode 100644 index 0000000000000000000000000000000000000000..8847b0b33bdc79f5479ff951791fd6bf9817a621 --- /dev/null +++ b/images/img_000094.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4e55e702297fea3a9478010ec627895ca241124396c3640a8940e874bebc0c83 +size 732776 diff --git a/images/img_000095.png b/images/img_000095.png new file mode 100644 index 0000000000000000000000000000000000000000..5ab445c76b9d85ca14401803564f962305326bd7 --- /dev/null +++ b/images/img_000095.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a230d07cc1a8bf7226bb978334b4f38f8b9bfbec67566f41c8f581dfe5b96bc5 +size 783351 diff --git a/images/img_000096.png b/images/img_000096.png new file mode 100644 index 0000000000000000000000000000000000000000..53f3a0b823c12fa01a9fbd682c3b3ee92ea39aed --- /dev/null +++ b/images/img_000096.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e7c56e2057206d50341a6fcd833748aa8c2ac8e67b5193c45f84697c1e410148 +size 732564 diff --git a/images/img_000097.png b/images/img_000097.png new file mode 100644 index 0000000000000000000000000000000000000000..75a974b767412fe729b51b1e40b9654cf919affd --- /dev/null +++ b/images/img_000097.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1cb327a27f524b45a2046ba01e43a9b80db43d25542a1ba20f1b73bfd6731e1d +size 769220 diff --git a/images/img_000098.png b/images/img_000098.png new file mode 100644 index 0000000000000000000000000000000000000000..78e027dc50dcdd76a390bdca6e5be26bf296e2de --- /dev/null +++ b/images/img_000098.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0fd04c39e4fb6de03ec79c8feb5b000fe36c5e58c14040b4ebd661c3ec61b5d5 +size 586251 diff --git a/images/img_000099.png b/images/img_000099.png new file mode 100644 index 0000000000000000000000000000000000000000..f010c47505755ec7f288dca062debbf050c68442 --- /dev/null +++ b/images/img_000099.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5e99c0ac689f83b25f1fc2e54a7b54c9f98366ee21d15979bf4de575211af419 +size 601021 diff --git a/images/img_000100.png b/images/img_000100.png new file mode 100644 index 0000000000000000000000000000000000000000..998672d62d6fea8756970f8cef49a5b9f170a5bf --- /dev/null +++ b/images/img_000100.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:34b9ba5309248d9fe490b0f018f9245e0df5788e94702ea46684e3ae4a2217bd +size 547146 diff --git a/images/img_000101.png b/images/img_000101.png new file mode 100644 index 0000000000000000000000000000000000000000..c8fc7f8af172533e01da0489f744447b0b84f256 --- /dev/null +++ b/images/img_000101.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8385cda6483172195a7bcf0ad4b082d33bbd81bb5e89c2328926abc1b9a318e5 +size 606784 diff --git a/images/img_000102.png b/images/img_000102.png new file mode 100644 index 0000000000000000000000000000000000000000..aa859ca31a86e7a1cc42785cc63f3f82fd8ee501 --- /dev/null +++ b/images/img_000102.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b19cc2ce8374303a67ec754e29b13b87920766f34ab6500a66b22b887ad18d13 +size 683715 diff --git a/images/img_000103.png b/images/img_000103.png new file mode 100644 index 0000000000000000000000000000000000000000..3ba86294b3b2b9c7bf2630009cc6af8d917fd7ae --- /dev/null +++ b/images/img_000103.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2cbbc0070bfdc1485f108130d63177933b2398bcc1fd4ed4baa828f115a2291a +size 768410 diff --git a/images/img_000104.png b/images/img_000104.png new file mode 100644 index 0000000000000000000000000000000000000000..01e6303c98741c9776f14b149873a0148c45fa6d --- /dev/null +++ b/images/img_000104.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c9b790c0e3b87398e72edb6942a374c0acb62cee53fefcb7a3f3f444065b7453 +size 744116 diff --git a/images/img_000105.png b/images/img_000105.png new file mode 100644 index 0000000000000000000000000000000000000000..a711dc6d0b5cee0b3374b02b9c2aba7a1c2e1873 --- /dev/null +++ b/images/img_000105.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6cd600887bce9f75269558e21406a0c17a0a918f54b4a31475cb74e4be44524c +size 763685 diff --git a/images/img_000106.png b/images/img_000106.png new file mode 100644 index 0000000000000000000000000000000000000000..f7771963dd323626fbd01351bf2c277dec31fd27 --- /dev/null +++ b/images/img_000106.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b595b6583877e5ccde929b7ecdf1080bc8e561d37c85d44c078e46f28be39edb +size 670651 diff --git a/images/img_000107.png b/images/img_000107.png new file mode 100644 index 0000000000000000000000000000000000000000..23f1a5763dba8415c3147cd5375b970a30c70ce9 --- /dev/null +++ b/images/img_000107.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:aacd4aa2f924b94989d1844547ae7ef4c4c15f8734bf42dbe4b000321ad9c061 +size 743760 diff --git a/images/img_000108.png b/images/img_000108.png new file mode 100644 index 0000000000000000000000000000000000000000..44adf53962cab14106e9bc05f302c69e80304f91 --- /dev/null +++ b/images/img_000108.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6342937ef57c6c7c04c885b7b6e817bfd7e3fb95170802cba0b50173f45f9d2c +size 774873 diff --git a/images/img_000109.png b/images/img_000109.png new file mode 100644 index 0000000000000000000000000000000000000000..5f01edb2ac29605f5da12cec1a6b92f2deab6938 --- /dev/null +++ b/images/img_000109.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c871908a0a623ab7faf8b6d70a0b9f27e45dce549201f50c9734b92bea0784a1 +size 663938 diff --git a/images/img_000110.png b/images/img_000110.png new file mode 100644 index 0000000000000000000000000000000000000000..90996659779e3927bd4519180b86662088bd5014 --- /dev/null +++ b/images/img_000110.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9cfdb5a756e0aa52dd3e40a3735d3870bc16c82da313507f2d73b26fe18b8aba +size 525740 diff --git a/images/img_000111.png b/images/img_000111.png new file mode 100644 index 0000000000000000000000000000000000000000..f6fb23f0923b3090dd272e00beb19c2923e1d638 --- /dev/null +++ b/images/img_000111.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1a26261144b41b30e245e7a01cb86c1965b35def5e873f3e62f147aba74a0c52 +size 504401 diff --git a/images/img_000112.png b/images/img_000112.png new file mode 100644 index 0000000000000000000000000000000000000000..3eea62bd9579940e1788cf0b52982769408a71df --- /dev/null +++ b/images/img_000112.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:26daf7905d8c2c3f9f2f740bd3b06f227609c7c23ab78d1f53608f3fad92300d +size 549932 diff --git a/images/img_000113.png b/images/img_000113.png new file mode 100644 index 0000000000000000000000000000000000000000..091823982c071c99b3f557bcd0f2fcc0dd4eeb0e --- /dev/null +++ b/images/img_000113.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:23ded4d7751066a8a0e27178b67d45b0529aa498e3c4798ef359557a3409b882 +size 553555 diff --git a/images/img_000114.png b/images/img_000114.png new file mode 100644 index 0000000000000000000000000000000000000000..29e47ac165acc9260ad3373114d2c560b4ba9160 --- /dev/null +++ b/images/img_000114.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:857546293f2c8c308b24fb199873072e5ddf860ebdee830f562f84bbe9b9e015 +size 703175 diff --git a/images/img_000115.png b/images/img_000115.png new file mode 100644 index 0000000000000000000000000000000000000000..9062c2ac3d18e83622e536ded1f42d71993194ed --- /dev/null +++ b/images/img_000115.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:178468171f028381b4284548ed073925cdd2abb3431116010b2e178891c28bcf +size 743163 diff --git a/images/img_000116.png b/images/img_000116.png new file mode 100644 index 0000000000000000000000000000000000000000..e33ca9d2e22fdfdb191de4815cd41cb75437c412 --- /dev/null +++ b/images/img_000116.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f32fa40f137fafce5a22c7c999d4f98566337c9e3465a752a6468fccbdbacdf5 +size 690040 diff --git a/images/img_000117.png b/images/img_000117.png new file mode 100644 index 0000000000000000000000000000000000000000..674a2e8f5f03c438645f131555f735befaa443fc --- /dev/null +++ b/images/img_000117.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7957d40babc29ba88146f77e6d727617c4cb474836877a6b0618bd3e08e1ba8a +size 727749 diff --git a/images/img_000118.png b/images/img_000118.png new file mode 100644 index 0000000000000000000000000000000000000000..34960978fcd021f4e93c862f1d24b95cd965377c --- /dev/null +++ b/images/img_000118.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5b5fbd77b4d94365e97dcdc67c4463916a606b26d50e51348defd30744e8e849 +size 647439 diff --git a/images/img_000119.png b/images/img_000119.png new file mode 100644 index 0000000000000000000000000000000000000000..e29f2cec9225ab406ca8f29a2f526c004a46d5fe --- /dev/null +++ b/images/img_000119.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f2e9fda04431218e043221e987b3c5889b49c459f969a6c05347cdd4d6c41360 +size 708279 diff --git a/images/img_000120.png b/images/img_000120.png new file mode 100644 index 0000000000000000000000000000000000000000..e27972771a78dadeac7d751870774c8ef863dfa0 --- /dev/null +++ b/images/img_000120.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fcc1a901d684da96c8f1f049fa3b9d51e65c957a111e881470f6a26fba0afbbf +size 660555 diff --git a/images/img_000121.png b/images/img_000121.png new file mode 100644 index 0000000000000000000000000000000000000000..4c48067a67ea188af4a925a2eaef2925125c634e --- /dev/null +++ b/images/img_000121.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8aed8ba74a500bad520f53f6e8c5fe252eb9b2899907a5fcf288620b8876eb5a +size 715623 diff --git a/images/img_000122.png b/images/img_000122.png new file mode 100644 index 0000000000000000000000000000000000000000..05f39666e9bf06eab1525baab7afb81d84827939 --- /dev/null +++ b/images/img_000122.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:724eaa5a15a557f65d506c1d96b122ab68e9e82e91bfc5fb7015f72a93920ffa +size 531413 diff --git a/images/img_000123.png b/images/img_000123.png new file mode 100644 index 0000000000000000000000000000000000000000..f4c764f6170bf88cf40d2ec8bd41c7c0122d1586 --- /dev/null +++ b/images/img_000123.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bb1b87bd62f1b71318622f8b0e83224da2851ec0bb8e6b8bec6517f32e17f518 +size 556333 diff --git a/images/img_000124.png b/images/img_000124.png new file mode 100644 index 0000000000000000000000000000000000000000..12ef0d51bf59793603745ecf0a612bf7dde10d83 --- /dev/null +++ b/images/img_000124.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1e710c6d219f518d44f190b6e25e6c1979c4f2ce227f4556e9ffb4354680682d +size 586977 diff --git a/images/img_000125.png b/images/img_000125.png new file mode 100644 index 0000000000000000000000000000000000000000..965dd71d8ce65f505b250f9c694bc69f71a8d2b3 --- /dev/null +++ b/images/img_000125.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:69bfb18a275000ce5ebf927770f8b369b9a20a74339750f1612dd26bae091986 +size 573608 diff --git a/images/img_000126.png b/images/img_000126.png new file mode 100644 index 0000000000000000000000000000000000000000..feab6a5ca5b17b3e744786fc0e84704b66695d41 --- /dev/null +++ b/images/img_000126.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:edc6b8498c3aec26f457e05dc0d9f1ec9cf66c566c2d023c535873f05bb33164 +size 813012 diff --git a/images/img_000127.png b/images/img_000127.png new file mode 100644 index 0000000000000000000000000000000000000000..14d8162d9579fb083903fa36954a66d2779ea0d3 --- /dev/null +++ b/images/img_000127.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:42a5240aac3f97fb6d6dec38e794e15954d42ff7b4b43dce29e739d3ffb2a604 +size 851265 diff --git a/images/img_000128.png b/images/img_000128.png new file mode 100644 index 0000000000000000000000000000000000000000..db52324db211e98f4c6b58e8f23aa0ee8d9e5739 --- /dev/null +++ b/images/img_000128.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:89b4a48de6063ff358d15b7e634c2aa83acd94daa650b40bdb6edbe9acfc536b +size 793119 diff --git a/images/img_000129.png b/images/img_000129.png new file mode 100644 index 0000000000000000000000000000000000000000..ff22af08fd589980c9c65b93563c566d2fd7382c --- /dev/null +++ b/images/img_000129.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ba544a3f9965938038d5621f108995e7b887d84b23265dcb6c093af3d0dd7054 +size 844847 diff --git a/images/img_000130.png b/images/img_000130.png new file mode 100644 index 0000000000000000000000000000000000000000..6998e129aed8aac23ec665aaea072f724fc4adea --- /dev/null +++ b/images/img_000130.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2a1cde47df490d01a9ab3637093e43c8c80ea70a33a21378712db6a8fdf29230 +size 798415 diff --git a/images/img_000131.png b/images/img_000131.png new file mode 100644 index 0000000000000000000000000000000000000000..9a51bfa8e56f98f9af4071ed19bc4a3db3a57ddc --- /dev/null +++ b/images/img_000131.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8b19ce4552c65f887edc5379c94f1d21c65d5cacad06281fcac6ca861c5a7313 +size 820452 diff --git a/images/img_000132.png b/images/img_000132.png new file mode 100644 index 0000000000000000000000000000000000000000..06abcb758de591de5fbfd350d7cf3abbe3833f51 --- /dev/null +++ b/images/img_000132.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:494424f0199704d3117d105b344017ddae378a30ccecd8434bd420350c19d5d4 +size 745124 diff --git a/images/img_000133.png b/images/img_000133.png new file mode 100644 index 0000000000000000000000000000000000000000..13cdefeffa2ec5c6c0bd993c72326186c4b3b0dd --- /dev/null +++ b/images/img_000133.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9c39a6e6eacc979d4fef0ecf2d344e6299fefe63946d8f840bee43e1c7128c59 +size 840667 diff --git a/images/img_000134.png b/images/img_000134.png new file mode 100644 index 0000000000000000000000000000000000000000..4410f928759156e54d5f04b18bb734ac23f495ed --- /dev/null +++ b/images/img_000134.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:411b698ae9b7f643806bc1b6470f34d6e129252ce8e536e580f67e6c496ef9e8 +size 653389 diff --git a/images/img_000135.png b/images/img_000135.png new file mode 100644 index 0000000000000000000000000000000000000000..c4c3afa1963119e81affe73782ab5c5a4e3cfe93 --- /dev/null +++ b/images/img_000135.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9c080749d0febd31dfa5279f1e0cb3454c4e516b41cf7f1eeeb2edaf0fdcdd4a +size 689457 diff --git a/images/img_000136.png b/images/img_000136.png new file mode 100644 index 0000000000000000000000000000000000000000..64769917fb1f71bab4dcc3b4acf5c67794a5de18 --- /dev/null +++ b/images/img_000136.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:00befb90240f02a031e661e50914021fadc326892c9eca11166cd4d975bf8e3d +size 706896 diff --git a/images/img_000137.png b/images/img_000137.png new file mode 100644 index 0000000000000000000000000000000000000000..c4e6f7f0b19332c92df4d39fe3280e4f720170cb --- /dev/null +++ b/images/img_000137.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1b1cf6d1b9bba07df5e0861b00dbd9adaac329e7040e8e77caffa3d32c1490cb +size 829868 diff --git a/images/img_000138.png b/images/img_000138.png new file mode 100644 index 0000000000000000000000000000000000000000..ca891c3b3337310467151c55dd16f652cad32146 --- /dev/null +++ b/images/img_000138.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:166fa39dac66456ed87a0ae4e28d3aba91be3f73855d5fa13a1334d990d625d0 +size 870340 diff --git a/images/img_000139.png b/images/img_000139.png new file mode 100644 index 0000000000000000000000000000000000000000..47354600bc7f26fc577676580e8b58c0f746c810 --- /dev/null +++ b/images/img_000139.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:15d3ae9051ab3867f1e3da63cce7d41c1d6c8a95a1dac98b68d65bffde068b66 +size 822550 diff --git a/images/img_000140.png b/images/img_000140.png new file mode 100644 index 0000000000000000000000000000000000000000..5c9e964c7480d3ac2c7359e8f0e90f02906f95c0 --- /dev/null +++ b/images/img_000140.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:652818e6257a3671fd3ce50769fc08fd5db28d77c860bbde78678ab0cf04c436 +size 842938 diff --git a/images/img_000141.png b/images/img_000141.png new file mode 100644 index 0000000000000000000000000000000000000000..2d342f8c9c49d426efaf2fa68105c497c9e7665d --- /dev/null +++ b/images/img_000141.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:20d437b99ad1529c1d699bdddef8187ad4558f8094a224d5aa596ef0a66ed41a +size 895782 diff --git a/images/img_000142.png b/images/img_000142.png new file mode 100644 index 0000000000000000000000000000000000000000..f421d25a58149b964ff3b50499b4fa24a4be2ba7 --- /dev/null +++ b/images/img_000142.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a6feaa680564dc22b3bb60570a51cf4ccc9578b83b6172494c2bf11acafb9c53 +size 851841 diff --git a/images/img_000143.png b/images/img_000143.png new file mode 100644 index 0000000000000000000000000000000000000000..e7df88c26d3ea2878f8124694fec50c0e080ddd6 --- /dev/null +++ b/images/img_000143.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d62fb387f75900f24a1d9c49299489b42e7da269a434bfa1430e87032d1f3af0 +size 823905 diff --git a/images/img_000144.png b/images/img_000144.png new file mode 100644 index 0000000000000000000000000000000000000000..e75894e50ca8fb3af0b28d0db16cd3973de08d89 --- /dev/null +++ b/images/img_000144.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a3249690a510ed1536f0b2c96a6bb69b5516f2bcf64c6e10aa5c432a5e0d6fa5 +size 859917 diff --git a/images/img_000145.png b/images/img_000145.png new file mode 100644 index 0000000000000000000000000000000000000000..13555a8e7c8d2342f1e56fb4eeb4b2e012e6d9ea --- /dev/null +++ b/images/img_000145.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0c385924ee9f5ace1f61be4fa841cf6668dcc6b439d79f4fe6319045a219f8df +size 679700 diff --git a/images/img_000146.png b/images/img_000146.png new file mode 100644 index 0000000000000000000000000000000000000000..311091993289fe17df656a191c72e9a4cb9d0191 --- /dev/null +++ b/images/img_000146.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:74de477a734e126de24514dc66f9c0ca6bd936ecc650c315668e39af25636045 +size 690651 diff --git a/images/img_000147.png b/images/img_000147.png new file mode 100644 index 0000000000000000000000000000000000000000..18b5b8ff28702e641fad21305427a45f11b631cc --- /dev/null +++ b/images/img_000147.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f56a9e9435825f17c6d0bc8dc445da28d58607f8d9c24de0ae535b7af51ce8ad +size 702892 diff --git a/images/img_000148.png b/images/img_000148.png new file mode 100644 index 0000000000000000000000000000000000000000..2505f49516b5dece87244074d3655c68740086be --- /dev/null +++ b/images/img_000148.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6ebed86a6726a6413b54885e28f8d30d82f0f4104f76628cdee1209191fdd748 +size 815311 diff --git a/images/img_000149.png b/images/img_000149.png new file mode 100644 index 0000000000000000000000000000000000000000..98dfd9fb3d143cbef219c9db945d1153a7ee199a --- /dev/null +++ b/images/img_000149.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:865fecffc974eb445d514eca9eb3b2e56c3b6106acbbc941a5fe7399fe22a905 +size 858755 diff --git a/images/img_000150.png b/images/img_000150.png new file mode 100644 index 0000000000000000000000000000000000000000..0c54bda6342c484b7fe67d0932c4ee850d9be28e --- /dev/null +++ b/images/img_000150.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:de848abcbec2220f17394768675bc082d1ef4e33772fb29df82798e8154f45ef +size 896427 diff --git a/images/img_000151.png b/images/img_000151.png new file mode 100644 index 0000000000000000000000000000000000000000..75578ea982f241d39aee11b9fbac74031b13b557 --- /dev/null +++ b/images/img_000151.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9c2c7089171e424ae5e5754a26136948f56feca06521e16f1a7629b17dd438ae +size 846210 diff --git a/images/img_000152.png b/images/img_000152.png new file mode 100644 index 0000000000000000000000000000000000000000..d3cdae59128c15505cb31c3d909c6f1c182dc1ff --- /dev/null +++ b/images/img_000152.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:da6e37a6ade6d5206a4b49fe9fb9588830387fa7ff8c8a8aa414670ff9ed58b5 +size 863806 diff --git a/images/img_000153.png b/images/img_000153.png new file mode 100644 index 0000000000000000000000000000000000000000..bcf7b9ac4f57bc526b3caef543bd8dcf49d3e318 --- /dev/null +++ b/images/img_000153.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fdb25743cfa7cbf71750de9d133d1e266360c63103adeace1e770ccb1151c015 +size 843978 diff --git a/images/img_000154.png b/images/img_000154.png new file mode 100644 index 0000000000000000000000000000000000000000..68d4e78c6749650d3429fe618023497619019d04 --- /dev/null +++ b/images/img_000154.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7a56a716182001451ea4a834a37d61da6c8b487d91fa252723dd441791764cbf +size 802712 diff --git a/images/img_000155.png b/images/img_000155.png new file mode 100644 index 0000000000000000000000000000000000000000..381dcfd30169d0e834d2c80ca57391524c1aaabc --- /dev/null +++ b/images/img_000155.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0d02b583e225fe64820bc11a4d2ebcf4ab8d0e6feb9a1d0905991bea602b7b97 +size 846186 diff --git a/images/img_000156.png b/images/img_000156.png new file mode 100644 index 0000000000000000000000000000000000000000..b5a4c0601a39b8cb4d8d1d6ef6e9b8cb0899dbf2 --- /dev/null +++ b/images/img_000156.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a67c734aaa1c9680540bf6c277696619aaad7e0dd83c3173b103feb1d9aedfc0 +size 839699 diff --git a/images/img_000157.png b/images/img_000157.png new file mode 100644 index 0000000000000000000000000000000000000000..5bec8a62f8769643ac91dd3b1bd8c0779bb64f94 --- /dev/null +++ b/images/img_000157.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9363b92fcbbbd1749a4192bf0ec65291544e2a86438a6780fd7a81907342af73 +size 711990 diff --git a/images/img_000158.png b/images/img_000158.png new file mode 100644 index 0000000000000000000000000000000000000000..92198dc2aa6a3acb411968244d87a13036100495 --- /dev/null +++ b/images/img_000158.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:98b45a89a92ca83eaa72f9be2491fa1e13e9872472ec3ffdf290a659b8afe4e2 +size 729294 diff --git a/images/img_000159.png b/images/img_000159.png new file mode 100644 index 0000000000000000000000000000000000000000..b5f287986c5b7f577bedb3cd79ef4dc94a1f49cb --- /dev/null +++ b/images/img_000159.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1d4ae26ae3fa66cb0af091339de44a730e28fe3ad94a6b94058a410132c6894c +size 718273 diff --git a/images/img_000160.png b/images/img_000160.png new file mode 100644 index 0000000000000000000000000000000000000000..c47a415d8d9ea82f0c3b5b35db05192d598b0007 --- /dev/null +++ b/images/img_000160.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:755c40209cfcb1249e938480485b9e3dc5b70f5d73b87368803f4ce926ca3a63 +size 811325 diff --git a/images/img_000161.png b/images/img_000161.png new file mode 100644 index 0000000000000000000000000000000000000000..06a7536f9c80614670c84632deeae679b1f70d1a --- /dev/null +++ b/images/img_000161.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cedff102f0d31d6bdf80761f413c976304cc22c4bac93d0609164abb76ef8a57 +size 866462 diff --git a/images/img_000162.png b/images/img_000162.png new file mode 100644 index 0000000000000000000000000000000000000000..e68de2acb62923e208d3e90ce5b388f6343e4fe2 --- /dev/null +++ b/images/img_000162.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f2229c8f772ea17be813972ebd5496000489d62dbea532ae511d673d4d08f17c +size 857099 diff --git a/images/img_000163.png b/images/img_000163.png new file mode 100644 index 0000000000000000000000000000000000000000..23ef1d048518c94dc69a0cdacad64488a770eaaa --- /dev/null +++ b/images/img_000163.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fad69701827a9f12bd56c1cfb54e2e9986b6878dda1d4017fa6cf9b968f4ada7 +size 829482 diff --git a/images/img_000164.png b/images/img_000164.png new file mode 100644 index 0000000000000000000000000000000000000000..fd22505c9f3dc8f0af2adc7c68af14df73c43148 --- /dev/null +++ b/images/img_000164.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:22c487c19f88b300f770b8645c5a9b01191edfb20ea9ae5dfcb2696d22fb0d16 +size 866446 diff --git a/images/img_000165.png b/images/img_000165.png new file mode 100644 index 0000000000000000000000000000000000000000..c623a699034bd9fd571af3aa33eaa859ed36bfae --- /dev/null +++ b/images/img_000165.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d15680f181c6c22d413c5cd7b0f48ef62a9dc0dac26c611d26ec81fde2009675 +size 874395 diff --git a/images/img_000166.png b/images/img_000166.png new file mode 100644 index 0000000000000000000000000000000000000000..3edb799c45dd6fecb2d49722162a27c326a4344c --- /dev/null +++ b/images/img_000166.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4aeb3e3bc87bdb2b745d7c83a1fc20f2e068d6741cd7b99fcd8e2f6ba8c3887b +size 844054 diff --git a/images/img_000167.png b/images/img_000167.png new file mode 100644 index 0000000000000000000000000000000000000000..f61f9da71c0d411ef2d3a1c7a82e43792818d18c --- /dev/null +++ b/images/img_000167.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cdc93624eccb11d445febcd5213a1d92fb41978378487163c22dfdbf2144c918 +size 614883 diff --git a/images/img_000168.png b/images/img_000168.png new file mode 100644 index 0000000000000000000000000000000000000000..7c03165240aac06a4be76f1c709a6ba279077cf5 --- /dev/null +++ b/images/img_000168.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:69286f786b2b0609830e0a0ddc9c925e891f9139e82ea5957d509096f4f110e0 +size 676050 diff --git a/images/img_000169.png b/images/img_000169.png new file mode 100644 index 0000000000000000000000000000000000000000..6bacc07404dc464cd26bdb158f5d2be00110071b --- /dev/null +++ b/images/img_000169.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:593c0a751bbdaea65a3ad0010154921f160b9b4222c2a8e939692f42cebdca25 +size 702972 diff --git a/images/img_000170.png b/images/img_000170.png new file mode 100644 index 0000000000000000000000000000000000000000..2e865d6568f5556dc186ce9947b874195bbbc53b --- /dev/null +++ b/images/img_000170.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:41708574cb00602b66c10a03890136a33153e377bdfb5ea7e1f6712a4cac037e +size 667882 diff --git a/images/img_000171.png b/images/img_000171.png new file mode 100644 index 0000000000000000000000000000000000000000..559c9589a04b1bee510961893b8a01393487bce3 --- /dev/null +++ b/images/img_000171.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4843308c6c40b4a1156872550ee7bd02b78bb0617718f68eba5948509ec0c401 +size 843427 diff --git a/images/img_000172.png b/images/img_000172.png new file mode 100644 index 0000000000000000000000000000000000000000..7889ce8eeaf1108d344258c2ac9b586b001e5fc0 --- /dev/null +++ b/images/img_000172.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7264fe9f5abd80a0b7c19e1248c40389ad6c631e9535811474f93669e510d06a +size 796076 diff --git a/images/img_000173.png b/images/img_000173.png new file mode 100644 index 0000000000000000000000000000000000000000..c83b9b2f97dc9bd6c22128af5e89c7b623101fc7 --- /dev/null +++ b/images/img_000173.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c806ca3d56c7284429831868d2e6e304324494562f9d3ffb93ad67b7a38b7462 +size 813285 diff --git a/images/img_000174.png b/images/img_000174.png new file mode 100644 index 0000000000000000000000000000000000000000..bf28c604eec46a45d67a0f3ed60cdcbd9e72bb8d --- /dev/null +++ b/images/img_000174.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b52060302de0e7caf49291847587e5d112843d1484100ac8533fc0a3798ba130 +size 785260 diff --git a/images/img_000175.png b/images/img_000175.png new file mode 100644 index 0000000000000000000000000000000000000000..0c57ceb6ba1fa645a1eacb44e031a81823c6a5d5 --- /dev/null +++ b/images/img_000175.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7ceacbfcd5653bd1e6b3b7be300a2dcb34466ddf885789cffd708c5906200194 +size 814384 diff --git a/images/img_000176.png b/images/img_000176.png new file mode 100644 index 0000000000000000000000000000000000000000..5587168241da773c05c6bc54315e51fd8426dba5 --- /dev/null +++ b/images/img_000176.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:69bb9b66553d6f4a2ea69ff2b621e3849b8c475c32075d9bfe1086f4ee88321a +size 764727 diff --git a/images/img_000177.png b/images/img_000177.png new file mode 100644 index 0000000000000000000000000000000000000000..4072f6e102afe3e0668e66924697fbb1783778e0 --- /dev/null +++ b/images/img_000177.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b11794a4e05ece8ac1a32adbadc55925e0def70bf1afe3e50a072e91b8d8ebb3 +size 829350 diff --git a/images/img_000178.png b/images/img_000178.png new file mode 100644 index 0000000000000000000000000000000000000000..5ab5adc519632e164319cfffe0a0003abd3866d6 --- /dev/null +++ b/images/img_000178.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0d475843b495f6ab01ba40c3c736ae09eee325830cf265d29e3ac78220010f38 +size 642408 diff --git a/images/img_000179.png b/images/img_000179.png new file mode 100644 index 0000000000000000000000000000000000000000..9882117af07f1aa77b82bae0be167e9749fc3a2d --- /dev/null +++ b/images/img_000179.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b455ae3b2ca4a678543996b73dfe087d9d7c5a75e7d4a988011af2af9a7d0026 +size 687858 diff --git a/images/img_000180.png b/images/img_000180.png new file mode 100644 index 0000000000000000000000000000000000000000..d7964abac1ad194ce72dddf074a8036040c3d640 --- /dev/null +++ b/images/img_000180.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f1ae236d975fbda5502736b15f3b1804794688efb1dbde52b2fe07d32d801d65 +size 654745 diff --git a/images/img_000181.png b/images/img_000181.png new file mode 100644 index 0000000000000000000000000000000000000000..5a3c523a6e124b5a917eef5a506a9bb8cbc3cd43 --- /dev/null +++ b/images/img_000181.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ab5b3433897fb8ab393d8c2be02c4f7cbef9e311bff0624e64d553d02bc94687 +size 765453 diff --git a/images/img_000182.png b/images/img_000182.png new file mode 100644 index 0000000000000000000000000000000000000000..c4d38db76cdb61ed65f489ade75cdbd5e2a62f62 --- /dev/null +++ b/images/img_000182.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b3e93d1db9e605bcce7dd9d60ab6f37a99d5c1d6fbfd941d96582a4c58618409 +size 812144 diff --git a/images/img_000183.png b/images/img_000183.png new file mode 100644 index 0000000000000000000000000000000000000000..978d509b8389bea59688f5c185520bdcab26d943 --- /dev/null +++ b/images/img_000183.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dedcd987d16abcfd92198ebb12dd2ecdd79209bc1fcecc6fbeca40cee9cf1d3d +size 780392 diff --git a/images/img_000184.png b/images/img_000184.png new file mode 100644 index 0000000000000000000000000000000000000000..a90fced5b7b8efbb69a26242db27de5c6dd29ed3 --- /dev/null +++ b/images/img_000184.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e27c61b897f6d64f34de8a2bbdc24c25fdbce511d8a22519ebbe30f482db770b +size 821084 diff --git a/images/img_000185.png b/images/img_000185.png new file mode 100644 index 0000000000000000000000000000000000000000..a47e27fbd99d161028bda214da6227e94b718987 --- /dev/null +++ b/images/img_000185.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fcd718e1b45865aedbaf7bf2852c033918e4d90e51dfe2e293796f7b68681905 +size 770339 diff --git a/images/img_000186.png b/images/img_000186.png new file mode 100644 index 0000000000000000000000000000000000000000..f10c63f3a14a384dd8a063b443ab305e13fb0824 --- /dev/null +++ b/images/img_000186.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c56b33fb9a2976251ffbb9d0fe9d6a4792bdd523429f7bb3e131843263ee14ef +size 807850 diff --git a/images/img_000187.png b/images/img_000187.png new file mode 100644 index 0000000000000000000000000000000000000000..f02c1b403414a348802377fae95e92f9b3a72106 --- /dev/null +++ b/images/img_000187.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6981543289240785ee70d9444d4499347800d340767a1e3116b15ae2c0c1e4f9 +size 780526 diff --git a/images/img_000188.png b/images/img_000188.png new file mode 100644 index 0000000000000000000000000000000000000000..9b77ec8f4afa94faf6c947028c8258a0d9839c0d --- /dev/null +++ b/images/img_000188.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b4d183cf71bce19e1a2da0d10d76a61120293e650ceba6d1c4bdbd308c530130 +size 815565 diff --git a/images/img_000189.png b/images/img_000189.png new file mode 100644 index 0000000000000000000000000000000000000000..8e9f52f3ea0ee331ee0bf44956db11ff0306bda7 --- /dev/null +++ b/images/img_000189.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4af96595dbc2dd1b877b02b68f71611b03aed720f118c4c1e2ac879e1df828e9 +size 783699 diff --git a/images/img_000190.png b/images/img_000190.png new file mode 100644 index 0000000000000000000000000000000000000000..a251eedae28c520599db4a57b0eae0d91f2bc3e0 --- /dev/null +++ b/images/img_000190.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2d3ab622ac8c18fe71abd147336a28e7006e5377aa666c4c363b01b02d0004b9 +size 465871 diff --git a/images/img_000191.png b/images/img_000191.png new file mode 100644 index 0000000000000000000000000000000000000000..28c0e42613d2849f30b5f1ea4ba2184739fabfe4 --- /dev/null +++ b/images/img_000191.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e216f1b181d7e544fd5d595b7ad0ce2f15211bfe002445edba4c928d7684ded0 +size 465421 diff --git a/images/img_000192.png b/images/img_000192.png new file mode 100644 index 0000000000000000000000000000000000000000..89104d1556fa54e20f94485982fdae782a436ea9 --- /dev/null +++ b/images/img_000192.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:89334c634303f19f05254703dabb467d7af699ab5e076e86e72945d253ed8dd6 +size 442495 diff --git a/images/img_000193.png b/images/img_000193.png new file mode 100644 index 0000000000000000000000000000000000000000..5a1d21f28c8a95fc3e87af7926821cfadd8b9f12 --- /dev/null +++ b/images/img_000193.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:71c2bce13f0247c4f32e50700f19ed0e827f52a71cfa30db7c30872b1539fffa +size 444745 diff --git a/images/img_000194.png b/images/img_000194.png new file mode 100644 index 0000000000000000000000000000000000000000..da8d82dc06d9fc5c641a1c04fc9c9896b6bfd28e --- /dev/null +++ b/images/img_000194.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cf7e923cc991b72d6e7c811bfcc9a61e1b94f69b9fc9b27b694e041563f76461 +size 440071 diff --git a/images/img_000195.png b/images/img_000195.png new file mode 100644 index 0000000000000000000000000000000000000000..e48f8290074340ad811aed87f1add8b087d4880e --- /dev/null +++ b/images/img_000195.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4644c8e963b9a8d04d4fc63172e1fa6176c8ed264baa07b7a9d90780dda52b1e +size 443129 diff --git a/images/img_000196.png b/images/img_000196.png new file mode 100644 index 0000000000000000000000000000000000000000..dc9cbd2f143da4ad49dc9df6f2b1702f34b4fefc --- /dev/null +++ b/images/img_000196.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f4a60355f4ab53391ea3fdd4cfeac2a978f46c5417340bc661daaf0c9e55381e +size 461793 diff --git a/images/img_000197.png b/images/img_000197.png new file mode 100644 index 0000000000000000000000000000000000000000..b83f3f822dc0b9f15d59bc1d2bd819a761453f55 --- /dev/null +++ b/images/img_000197.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:543dcda4d9307021190bce7263c3e69b7021f3740820707b72c0105a6702651e +size 462061 diff --git a/images/img_000198.png b/images/img_000198.png new file mode 100644 index 0000000000000000000000000000000000000000..f7b70cedf8d6ea4ef451b3041c8d99cdc1599971 --- /dev/null +++ b/images/img_000198.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9fd404adce79f5c872e76eea8a497e648fc1e48242493ff9ff5b8d0c89377824 +size 450693 diff --git a/images/img_000199.png b/images/img_000199.png new file mode 100644 index 0000000000000000000000000000000000000000..2fb9b523e48e6df287ae54e6ed02b1dc31bdd511 --- /dev/null +++ b/images/img_000199.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c443be18a2a9f9d604974d42d09f0d072d1390a0d1aa21ef25b90fa9c5cd7abb +size 446531 diff --git a/images/img_000200.png b/images/img_000200.png new file mode 100644 index 0000000000000000000000000000000000000000..1f28cd13f90a3471ee07f3b071cbe43eb97a918f --- /dev/null +++ b/images/img_000200.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:79afe697383fcdc74084b7b772b890bbee1023895d1bd51fdf8de558eff67ff7 +size 442426 diff --git a/images/img_000201.png b/images/img_000201.png new file mode 100644 index 0000000000000000000000000000000000000000..27fcc654fce2323b10e6e43574e23e684b04e397 --- /dev/null +++ b/images/img_000201.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:02897688902a0932ec6ec8557c2adefde77b5350632a5fe21c8ffd1186892a54 +size 442513 diff --git a/images/img_000202.png b/images/img_000202.png new file mode 100644 index 0000000000000000000000000000000000000000..0957cf3ae8bd7286030bb19f7aecfc2c45d4be08 --- /dev/null +++ b/images/img_000202.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:45691b9f7a864995d0a0affeb9dfc073d369b664ea850dc13c8e9783a9fbf7df +size 433319 diff --git a/images/img_000203.png b/images/img_000203.png new file mode 100644 index 0000000000000000000000000000000000000000..1fdf877cf97e3c2677dfc631129782d2aa6de737 --- /dev/null +++ b/images/img_000203.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:562e732a84ad74eb00f5b0d1638acc33c2d9e8f071a2f8c8a69c51d361c35f11 +size 447484 diff --git a/images/img_000204.png b/images/img_000204.png new file mode 100644 index 0000000000000000000000000000000000000000..f7b6bb65ff9df3abcabc3589b643f7d83532aabb --- /dev/null +++ b/images/img_000204.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:088ea29e90fbdfbd6218a1a4ed89dc2658f334fe6e244837cce9905794d843cd +size 511130 diff --git a/images/img_000205.png b/images/img_000205.png new file mode 100644 index 0000000000000000000000000000000000000000..74f04fffee26cca859dd106e40e344d3964293c9 --- /dev/null +++ b/images/img_000205.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:185d1d2f77fd1b1693142f4c7c4aa61c125278784434d6ebb91e567165c7328d +size 498395 diff --git a/images/img_000206.png b/images/img_000206.png new file mode 100644 index 0000000000000000000000000000000000000000..345f4cecc79e847e5e40e4af11bd66413a46c5fb --- /dev/null +++ b/images/img_000206.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4d0b5d7323a5b498ce6e4181edb26922f6b38a2164f2e07c7035455aa4ffebfc +size 498688 diff --git a/images/img_000207.png b/images/img_000207.png new file mode 100644 index 0000000000000000000000000000000000000000..712bbb5bacde4ba139887edbe296b3683e15a149 --- /dev/null +++ b/images/img_000207.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:afd78c1757192d3a18588906f0affdd6b53ff8882525b5bb718e752695d2241a +size 501949 diff --git a/images/img_000208.png b/images/img_000208.png new file mode 100644 index 0000000000000000000000000000000000000000..63963ac2ea1b424f13c4b6df04e278fc16722a8e --- /dev/null +++ b/images/img_000208.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4fca070dc34625fb66ffa77eb1413f3b57ed8abaebccefcf93230c869c43665e +size 518356 diff --git a/images/img_000209.png b/images/img_000209.png new file mode 100644 index 0000000000000000000000000000000000000000..ad8690abe6887f52beb28df55994d5b170a0c357 --- /dev/null +++ b/images/img_000209.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7002a6b9082188959b18e172800ef0e3cb91222abf0a5c46e1d0ac7df54eddd2 +size 517345 diff --git a/images/img_000210.png b/images/img_000210.png new file mode 100644 index 0000000000000000000000000000000000000000..00e8517f4984a58bc817b3a57eb78e4a8ab48f00 --- /dev/null +++ b/images/img_000210.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:470aaeee93c3eded69646c6dc9b158f2a3ce6f3704e8c3083c2cd17a38ac701b +size 506958 diff --git a/images/img_000211.png b/images/img_000211.png new file mode 100644 index 0000000000000000000000000000000000000000..83cd507370419669f8640e655ad0a11079f44f5b --- /dev/null +++ b/images/img_000211.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d91b528dbc9327f685904b9d6bcf3fc93d340ea9ae157245a9c87f0074055e75 +size 489421 diff --git a/images/img_000212.png b/images/img_000212.png new file mode 100644 index 0000000000000000000000000000000000000000..90e3971884968f05e5806ccd0cd0ab3dd4ff2a47 --- /dev/null +++ b/images/img_000212.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:488f2350ba655f446cb5c284dfaf204c598b360c116b7fda4b834ea08bc7ebd8 +size 499947 diff --git a/images/img_000213.png b/images/img_000213.png new file mode 100644 index 0000000000000000000000000000000000000000..40809a5b547d7f6eaa558ae412757465e27d7a54 --- /dev/null +++ b/images/img_000213.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:454f16d046f8be7181a147b08878f9325c5f8f59e8ccabb9de6efbdf2957179b +size 490832 diff --git a/images/img_000214.png b/images/img_000214.png new file mode 100644 index 0000000000000000000000000000000000000000..aab2055b65f0359b7a92bb5d4b809cf31f86f7ca --- /dev/null +++ b/images/img_000214.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2a42045d04b9206dd7743582582a9a0e88c38d117b40301c9f4e4f8409f02db0 +size 533533 diff --git a/images/img_000215.png b/images/img_000215.png new file mode 100644 index 0000000000000000000000000000000000000000..a6a05daadf65524e7c37a9bad171f63a0d116914 --- /dev/null +++ b/images/img_000215.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4a1c4e2c1f661fe6ae2610705827faa6f1a5e76dee823b17c4c18a9e74d97dd9 +size 518519 diff --git a/images/img_000216.png b/images/img_000216.png new file mode 100644 index 0000000000000000000000000000000000000000..23baf9d8a8ee745956f98899908e01f642cd82f4 --- /dev/null +++ b/images/img_000216.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e19fbaf3bba3d82f50e532a04038a48316e0508d8f50b02070e08cdc3a0e7f42 +size 527115 diff --git a/images/img_000217.png b/images/img_000217.png new file mode 100644 index 0000000000000000000000000000000000000000..c0261fe642ffcd352e9703955bde609d7fd3f061 --- /dev/null +++ b/images/img_000217.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9e1d540b6575d5b2cbbf4a4c46810e5d876576d96cb96d2969e35a898979980d +size 512967 diff --git a/images/img_000218.png b/images/img_000218.png new file mode 100644 index 0000000000000000000000000000000000000000..dbbb2f3a3cc0f70d06e94245e054b4fd3dc15733 --- /dev/null +++ b/images/img_000218.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:65e47f69a1329bbc84dbfbce56986626c1465082bd1ce2a6bdf84214e1b9e0df +size 507337 diff --git a/images/img_000219.png b/images/img_000219.png new file mode 100644 index 0000000000000000000000000000000000000000..792db124553631ef7072fda520ccd50bf874d9fa --- /dev/null +++ b/images/img_000219.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f00a51ed187018dee86499d2793d5589869c6c3ef83852b74d73acc34a1cd04e +size 527025 diff --git a/images/img_000220.png b/images/img_000220.png new file mode 100644 index 0000000000000000000000000000000000000000..c2b41a1db09d024c2d5385be006648264a79a919 --- /dev/null +++ b/images/img_000220.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:509dd53c84967ec1c8f0e28139177d49661bd4a3f7f0e911738aa546b0e6aeca +size 481260 diff --git a/images/img_000221.png b/images/img_000221.png new file mode 100644 index 0000000000000000000000000000000000000000..d54958d4c74a5e7733fcb4517cd9202b57f84711 --- /dev/null +++ b/images/img_000221.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:beafa571c221ce3f0ab8aad9ffb54e9af4cea52ad67ee2e1e7a3bb119128f39a +size 520769 diff --git a/images/img_000222.png b/images/img_000222.png new file mode 100644 index 0000000000000000000000000000000000000000..9ff290920e52814b8a91a84969964609a5650229 --- /dev/null +++ b/images/img_000222.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:76dd20d9efc7c57789636d2f1296b4ad7148236a93b77fb839c59c5b768a24f3 +size 523574 diff --git a/images/img_000223.png b/images/img_000223.png new file mode 100644 index 0000000000000000000000000000000000000000..c08c349b3f762078ddae6e5abc2cc01f56753470 --- /dev/null +++ b/images/img_000223.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4bfed6e438654c996a203794a875b1fa4b66666bf0a2281fabb57288d4c43e29 +size 511916 diff --git a/images/img_000224.png b/images/img_000224.png new file mode 100644 index 0000000000000000000000000000000000000000..8dff4bc47c1c8694858083b47eb4fa50f91a9e66 --- /dev/null +++ b/images/img_000224.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f2f4c63b1e917281c49f53f556abee3cc78908d94ac2a6cac5a436df75b0cb72 +size 536047 diff --git a/images/img_000225.png b/images/img_000225.png new file mode 100644 index 0000000000000000000000000000000000000000..52aa9de1cd3bbe7fa3a40eff02fabeb3efe96cc9 --- /dev/null +++ b/images/img_000225.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9d0bd3313dadb6aead7d825b12584b316ac93bb0e28bf58771c4dd06fd26b216 +size 526797 diff --git a/images/img_000226.png b/images/img_000226.png new file mode 100644 index 0000000000000000000000000000000000000000..ffa4dd4c9685081702ab9f84f56c3087d69972b4 --- /dev/null +++ b/images/img_000226.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:00acdd16de805e499cd2bddb2f9110964e0576eaafae28985b733845c716c04f +size 482302 diff --git a/images/img_000227.png b/images/img_000227.png new file mode 100644 index 0000000000000000000000000000000000000000..d0e5078ae3f63bae7d38fc636393cceba2816952 --- /dev/null +++ b/images/img_000227.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:61a96f484be6eaa077d8f7a9458f847ff91e8a29305d89085e39221d8794f9ad +size 479938 diff --git a/images/img_000228.png b/images/img_000228.png new file mode 100644 index 0000000000000000000000000000000000000000..608aceca5dce1eef702b2ecceeb8a9095d28e25f --- /dev/null +++ b/images/img_000228.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f88cd713f85d13ea97857d4115385f029209be267840ecc98b84ff5c418b3b8f +size 513898 diff --git a/images/img_000229.png b/images/img_000229.png new file mode 100644 index 0000000000000000000000000000000000000000..b3aa329c5e384d162377f9d63d8441dc2173cc0d --- /dev/null +++ b/images/img_000229.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cee3c556718e88697f497d809e9b94a76f74f5c55f26a7392754ae4fc1354d61 +size 530706 diff --git a/images/img_000230.png b/images/img_000230.png new file mode 100644 index 0000000000000000000000000000000000000000..0afa27d0adb53036a424418e56d0d5ae978be203 --- /dev/null +++ b/images/img_000230.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3cb3611927cfd0c44b6fdb989b9825ae86a4fa3413f4fe235e8bec1d01083ccd +size 509225 diff --git a/images/img_000231.png b/images/img_000231.png new file mode 100644 index 0000000000000000000000000000000000000000..a3a86455149b42cdb565b7596cd9aae453f32fe4 --- /dev/null +++ b/images/img_000231.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:05898287f6f5165f63024c574a75a726bf53d5cfe6b1e84e194f80a20fb71b1d +size 520308 diff --git a/images/img_000232.png b/images/img_000232.png new file mode 100644 index 0000000000000000000000000000000000000000..518369083d3e81249d21f851d15b2767bfd804e8 --- /dev/null +++ b/images/img_000232.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7819289c44d2a816e2a9cebe8bf1d55e703d2ed26408c60c45f806ab0f7ab95b +size 466594 diff --git a/images/img_000233.png b/images/img_000233.png new file mode 100644 index 0000000000000000000000000000000000000000..0ad37ddb625744451977cc512b3e40c36e32832d --- /dev/null +++ b/images/img_000233.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bb3a59f1eb9cd5972322cb4f906a7ca0f7d964078874d6a342435a66370ba8e9 +size 475955 diff --git a/images/img_000234.png b/images/img_000234.png new file mode 100644 index 0000000000000000000000000000000000000000..9090f5a9a6d70d5b68be1c38a5b5d0c129eb4a0d --- /dev/null +++ b/images/img_000234.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:efc3e5e59e617d0090bdf7cc3ec538605dc5cb9ddcced6096d36b2f7e8f20136 +size 470109 diff --git a/images/img_000235.png b/images/img_000235.png new file mode 100644 index 0000000000000000000000000000000000000000..877cdca5d1b3a2a61062a057904cac5f39fe47bc --- /dev/null +++ b/images/img_000235.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fe1cb12fb4b6ded8bffa3ce8c6329b5827e9f81aa15f50d6b06d0eb68d88d8f0 +size 456839 diff --git a/images/img_000236.png b/images/img_000236.png new file mode 100644 index 0000000000000000000000000000000000000000..290a0c652cdf0f2c27398e9a9012b4c22736c18d --- /dev/null +++ b/images/img_000236.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1064e834e1b74c394802286bb01a4501e2a1c3242593417b2b889583825b3216 +size 443604 diff --git a/images/img_000237.png b/images/img_000237.png new file mode 100644 index 0000000000000000000000000000000000000000..089682558f1b1c4853a9d19b03d163627a0f3530 --- /dev/null +++ b/images/img_000237.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3f7af8567c3cfe4ecd0268bb09524380c8e281cc63ef76b89ba48c923eba9eef +size 439355 diff --git a/images/img_000238.png b/images/img_000238.png new file mode 100644 index 0000000000000000000000000000000000000000..51b4970a581a1b67fd403dd613fc9cc3b4cb29cd --- /dev/null +++ b/images/img_000238.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3711ac7f30f65c0695ac4b318c6447bc383837b82a70c562c048ca57228d5e6a +size 440157 diff --git a/images/img_000239.png b/images/img_000239.png new file mode 100644 index 0000000000000000000000000000000000000000..9f155c8dd7c51d3744a3893d7760dfcbf5533e8c --- /dev/null +++ b/images/img_000239.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:54a8b5bf4d093984790787d2aed8d0a99b9bc92d1d97dce0e0358c6c71f9078e +size 472245 diff --git a/images/img_000240.png b/images/img_000240.png new file mode 100644 index 0000000000000000000000000000000000000000..4fd50d4dd653796384da6e8d870783c9cdf4f200 --- /dev/null +++ b/images/img_000240.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5ba468dfd0f49fcda822e1160e798898b539fb6014e1c387cb5ae3f5c0a5f92d +size 467873 diff --git a/images/img_000241.png b/images/img_000241.png new file mode 100644 index 0000000000000000000000000000000000000000..4b61e03b8df470247dc33358ef4401fd1d49a26c --- /dev/null +++ b/images/img_000241.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fe8ba5a9377649be4aa09360c2c2aa1c08ef17a2c24f1351a09b9fcdde1167de +size 452659 diff --git a/images/img_000242.png b/images/img_000242.png new file mode 100644 index 0000000000000000000000000000000000000000..0d6c336942c91035d2be81b98159489acfe38aaa --- /dev/null +++ b/images/img_000242.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e9d28a69c9bb4a509efa93e8ca64ec47cb1e5f7014933cfb545f392843603ee2 +size 452305 diff --git a/images/img_000243.png b/images/img_000243.png new file mode 100644 index 0000000000000000000000000000000000000000..97352cb519d215b50b18442c167a42d45b260b08 --- /dev/null +++ b/images/img_000243.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e913491cf5ffa0ac5ec2e2e654a2eccb405a98caf329f21061761f8a80601a4b +size 451654 diff --git a/images/img_000244.png b/images/img_000244.png new file mode 100644 index 0000000000000000000000000000000000000000..a7d9f82de7365c81cef9ec07a7f35570d8cad044 --- /dev/null +++ b/images/img_000244.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:979dbf40d1d4371e2020c4ea1cd272fd1ac43161330a1e5d3333870e7b9b107b +size 456377 diff --git a/images/img_000245.png b/images/img_000245.png new file mode 100644 index 0000000000000000000000000000000000000000..8f20c0502aa1bac4fcfd461471472f5e410ebd79 --- /dev/null +++ b/images/img_000245.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c01dc0db41ccb97104e7d22e03f7e03dc1f609c195475498a0ddf30fba72af0e +size 443571 diff --git a/images/img_000246.png b/images/img_000246.png new file mode 100644 index 0000000000000000000000000000000000000000..88e199d53c9076aa7fa0058403a0d75d599d02f6 --- /dev/null +++ b/images/img_000246.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:59546184c04d2899e9b4f9afae634069056de15b8b7b44ec45774087c895619f +size 462611 diff --git a/images/img_000247.png b/images/img_000247.png new file mode 100644 index 0000000000000000000000000000000000000000..2cca3fc668343023dad6c020a3d814d7f17be300 --- /dev/null +++ b/images/img_000247.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:808c0e39c2ee780fad55a2c21bde7ee8c3f752c1b5736c794ef198671c4ae75b +size 502042