Datasets:
Tasks:
Image-to-3D
Sub-tasks:
semantic-segmentation
Languages:
English
ArXiv:
Tags:
3d scene understanding
3d-scene-completion
aerial perception
autonomous flying
dataset
benchmark
License:
Upload README.md with huggingface_hub
Browse files
README.md
CHANGED
|
@@ -1,3 +1,81 @@
|
|
| 1 |
-
|
| 2 |
-
|
| 3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# OccuFly Dataset
|
| 2 |
+
|
| 3 |
+
A 3D Vision Benchmark for Semantic Scene Completion from the Aerial Perspective. **CVPR 2026 Oral**.
|
| 4 |
+
|
| 5 |
+
📄 **Paper**: [arXiv:2512.20770](https://arxiv.org/abs/2512.20770)
|
| 6 |
+
🌐 **Project Page**: [markus-42.github.io/publications/2026/occufly/](https://markus-42.github.io/publications/2026/occufly/)
|
| 7 |
+
📚 **Full Documentation**: See [https://github.com/markus-42/occufly](https://github.com/markus-42/occufly)
|
| 8 |
+
|
| 9 |
+
---
|
| 10 |
+
|
| 11 |
+
## Quick Start
|
| 12 |
+
|
| 13 |
+
### 1. Install Dependencies
|
| 14 |
+
|
| 15 |
+
```bash
|
| 16 |
+
pip install huggingface-hub tqdm numpy Pillow
|
| 17 |
+
```
|
| 18 |
+
|
| 19 |
+
### 2. Download Dataset
|
| 20 |
+
|
| 21 |
+
Use the download script located in this repository:
|
| 22 |
+
|
| 23 |
+
```bash
|
| 24 |
+
# Download all scenes
|
| 25 |
+
python download_occufly.py
|
| 26 |
+
|
| 27 |
+
# Download specific split
|
| 28 |
+
python download_occufly.py --split train
|
| 29 |
+
python download_occufly.py --split validation
|
| 30 |
+
python download_occufly.py --split test
|
| 31 |
+
|
| 32 |
+
# Download specific scenes (1-9)
|
| 33 |
+
python download_occufly.py --scenes 1 2 3
|
| 34 |
+
|
| 35 |
+
# Include predicted depth maps
|
| 36 |
+
python download_occufly.py --include_depth_predictions
|
| 37 |
+
python download_occufly.py --split train --include_depth_predictions
|
| 38 |
+
|
| 39 |
+
# Download only predicted depth maps
|
| 40 |
+
python download_occufly.py --only_depth_predictions
|
| 41 |
+
|
| 42 |
+
# Custom output directory
|
| 43 |
+
python download_occufly.py --output ./my_data
|
| 44 |
+
```
|
| 45 |
+
|
| 46 |
+
## Dataset Splits
|
| 47 |
+
|
| 48 |
+
- **Train**: Scenes 1-5
|
| 49 |
+
- **Validation**: Scenes 6-7
|
| 50 |
+
- **Test**: Scenes 8-9
|
| 51 |
+
|
| 52 |
+
Each scene has 3 altitudes: 30m, 40m, 50m
|
| 53 |
+
|
| 54 |
+
|
| 55 |
+
## Citation
|
| 56 |
+
|
| 57 |
+
```bibtex
|
| 58 |
+
@misc{gross2025occufly,
|
| 59 |
+
title={{OccuFly}: A 3D Vision Benchmark for Semantic Scene Completion from the Aerial Perspective},
|
| 60 |
+
author={Markus Gross and Sai B. Matha and Aya Fahmy and Rui Song and Daniel Cremers and Henri Meess},
|
| 61 |
+
year={2025},
|
| 62 |
+
eprint={2512.20770},
|
| 63 |
+
archivePrefix={arXiv},
|
| 64 |
+
primaryClass={cs.CV},
|
| 65 |
+
url={https://arxiv.org/abs/2512.20770}
|
| 66 |
+
}
|
| 67 |
+
```
|
| 68 |
+
|
| 69 |
+
## License
|
| 70 |
+
|
| 71 |
+
This dataset is licensed under the **Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International** (CC BY-NC-SA 4.0) license.
|
| 72 |
+
|
| 73 |
+
- **You can**: Share and adapt the material for non-commercial purposes
|
| 74 |
+
- **You must**: Give credit, indicate changes, and use the same license for derivative works
|
| 75 |
+
- **You cannot**: Use the material for commercial purposes without permission
|
| 76 |
+
|
| 77 |
+
For full license details, see [creativecommons.org/licenses/by-nc-sa/4.0/](https://creativecommons.org/licenses/by-nc-sa/4.0/)
|
| 78 |
+
|
| 79 |
+
## Documentation
|
| 80 |
+
|
| 81 |
+
For detailed documentation, visit the [https://github.com/markus-42/occufly](https://github.com/markus-42/occufly).
|