Dataset Viewer
The dataset viewer is not available for this subset.
Cannot get the split names for the config 'default' of the dataset.
Exception: SplitsNotFoundError
Message: The split names could not be parsed from the dataset config.
Traceback: Traceback (most recent call last):
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 290, in _generate_tables
pa_table = paj.read_json(
io.BytesIO(batch), read_options=paj.ReadOptions(block_size=block_size)
)
File "pyarrow/_json.pyx", line 342, in pyarrow._json.read_json
File "pyarrow/error.pxi", line 155, in pyarrow.lib.pyarrow_internal_check_status
return check_status(status)
File "pyarrow/error.pxi", line 92, in pyarrow.lib.check_status
raise convert_status(status)
pyarrow.lib.ArrowInvalid: JSON parse error: Column() changed from object to string in row 0
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 286, in get_dataset_config_info
for split_generator in builder._split_generators(
~~~~~~~~~~~~~~~~~~~~~~~~~^
StreamingDownloadManager(base_path=builder.base_path, download_config=download_config)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
)
^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 101, in _split_generators
pa_table = next(iter(self._generate_tables(**splits[0].gen_kwargs, allow_full_read=False)))[1]
~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 304, in _generate_tables
batch = json_encode_fields_in_json_lines(original_batch, json_field_paths)
File "/usr/local/lib/python3.14/site-packages/datasets/utils/json.py", line 111, in json_encode_fields_in_json_lines
examples = [ujson_loads(line) for line in original_batch.splitlines()]
~~~~~~~~~~~^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/utils/json.py", line 20, in ujson_loads
return pd.io.json.ujson_loads(*args, **kwargs)
~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^
ValueError: Expected object or value
The above exception was the direct cause of the following exception:
Traceback (most recent call last):
File "/src/services/worker/src/worker/job_runners/config/split_names.py", line 68, in compute_split_names_from_streaming_response
for split in get_dataset_split_names(
~~~~~~~~~~~~~~~~~~~~~~~^
path=dataset,
^^^^^^^^^^^^^
config_name=config,
^^^^^^^^^^^^^^^^^^^
token=hf_token,
^^^^^^^^^^^^^^^
)
^
File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 340, in get_dataset_split_names
info = get_dataset_config_info(
path,
...<6 lines>...
**config_kwargs,
)
File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 291, in get_dataset_config_info
raise SplitsNotFoundError("The split names could not be parsed from the dataset config.") from err
datasets.inspect.SplitsNotFoundError: The split names could not be parsed from the dataset config.Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
SigLIP 2 visual tokens for HICO-DET (UMI-HOI)
Precomputed SigLIP 2 (google/siglip2-giant-opt-patch16-384, not the first-generation
SigLIP) image tokens for every HICO-DET image, used as the vision-language input of UMI-HOI (Unifying Multimodal Information with Semantic Multi-Head Attention for
Human-Object Interaction Detection, CVPR Findings 2026). Code:
https://github.com/ga94zot/UMI-HOI
What is in each file
| Encoder | google/siglip2-giant-opt-patch16-384, default AutoProcessor preprocessing (384x384, patch 16) |
| Per image | Tensor(577, 1536) float32: rows 0..575 = 24x24 patch tokens (last_hidden_state), row 576 = pooled image token (get_image_features pooled output) |
| Images | HICO-DET train2015 (38,118) and test2015 (9,658), keyed by image stem, e.g. HICO_train2015_00000001 |
| Storage | <split>/shard-XXXX.safetensors (~10 GiB each, bit-exact fp32), <split>/index.json (stem -> shard), shards.json (sha256, sizes) |
The features were generated with generate_feature.py (included) by the UMI-HOI authors.
Loading
Restore the per-image .pt layout that the UMI-HOI data loader reads:
python hf_download_features.py --repo-id <this repo> --out /data/hico_siglip_feature
python main.py ... --llava-token-path /data/hico_siglip_feature
Or read a shard directly:
from safetensors import safe_open
with safe_open("train/shard-0000.safetensors", framework="pt") as f:
tokens = f.get_tensor("HICO_train2015_00000001") # (577, 1536)
patches, pooled = tokens[:-1], tokens[-1]
License
The tensors are derived from HICO-DET images (Chao et al., research use only; see the HICO-DET terms) with an Apache-2.0 model. They are provided for research use under the same terms as HICO-DET.
Citation
@inproceedings{wu2026umihoi,
title = {Unifying Multimodal Information with Semantic Multi-Head Attention for Human-Object Interaction Detection},
author = {Wu, Yuankai and others},
booktitle = {CVPR Findings},
year = {2026}
}
- Downloads last month
- 126