Datasets:
Formats:
parquet
Languages:
English
Size:
10K - 100K
ArXiv:
Tags:
video-understanding
multi-evidence-reasoning
long-video
temporal-reasoning
spatial-reasoning
video-qa
License:
Commit History
-
Fix CVPR 2026 badge link to CVF open-access paper URL
Verified signature30b42e9 -
Add Lite-v2 (manually refined) config + README note; CVPR 2026 acceptance
Verified signature41206e1
-
Merge branch 'main' of https://huggingface.co/datasets/DanBenAmi/HERBench
e5ae45b -
Update data files and README with latest annotations
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>0b4b682
-
Fix 2>&1 bug in tar archive creation script
Verified signature289280c -
Add troubleshooting section to README
Verified signaturea28439b -
Fix corrupted tar archives (2>&1 bug in creation script)
Verified signature4187094
-
Merge branch 'main' of https://huggingface.co/datasets/DanBenAmi/HERBench
de0e0c7 -
Update dataset card with accurate statistics and links
- Update title to match paper - Update arXiv citation to 2512.14870 - Update project page to gabrieleserussi.github.io/HERBench/ - Add accurate statistics for Full (27,936 questions, 335 videos, 424s avg) and Lite (5,960 questions, 68 videos, 421s avg) - Add MRFS Analysis section with comparison chart - Remove coming soon placeholders from linksc174cf2
-
Upload folder using huggingface_hub
Verified signature828cefc -
Reorder questions by task_type to match README order
Reordered all questions in both Full and Lite versions to follow the task order shown in README.md: 1. Temporal Shot Ordering (TSO) 2. Multi-Person Duration Reasoning (MPDR) 3. Action Sequence Integrity Identification (ASII) 4. Appearance-Grounded Behavior & Interactions (AGBI) 5. Appearance-Grounded Attribute Recognition (AGAR) 6. Appearance-Grounded Localization & Trajectory (AGLT) 7. False Action Memory (FAM) 8. Scene Verification & Arrangement (SVA) 9. False Object Memory (FOM) 10. Multi-Entities Grounding and Localization (MEGL) 11. Action Counting (AC) 12. Region Localized People Counting (RLPC) This makes the dataset browsing more intuitive and aligns with the task family groupings in the documentation. Changes: - Reordered data/herbench_annotations.json (27,936 questions) - Reordered data/herbench_annotations_lite.json (5,960 questions) - Regenerated data/herbench_full.parquet - Regenerated data/herbench_lite.parquet8247391
-
Convert to Parquet format for Dataset Viewer compatibility
HuggingFace deprecated custom loading scripts. Switching to Parquet format with auto-loading: Changes: - Remove deprecated scripts: HERBench.py, herbench.py - Add Parquet files with Git LFS: herbench_full.parquet (4.87 MB), herbench_lite.parquet (1.02 MB) - Update README.md YAML to use Parquet with explicit configs - Add comprehensive data format documentation and usage examples - Update .gitattributes to track Parquet files with Git LFS - Keep original JSON files for users who need raw format Parquet format ensures: - Reliable schema handling (fixes Arrow conversion errors) - Better performance for large datasets - Full compatibility with HF Dataset Viewer - Support for varying metadata structures across task types471524a
-
Remove explicit configs from YAML to let HF auto-detect HERBench.py
Explicitly declaring configs without data_files was causing HF to fall back to JSON auto-loading. By removing the configs section, HF should auto-detect and use the HERBench.py loading script.58bf035 -
Add HERBench.py (capitalized) for HF dataset loader
HuggingFace expects the loading script filename to match the dataset name. Since the repo is DanBenAmi/HERBench, the script should be HERBench.py. Keeping both herbench.py (lowercase) and HERBench.py (capitalized) for compatibility.d5b91b2 -
Add configs to YAML to force herbench.py usage
Without explicit config names in the YAML frontmatter, HF Dataset Viewer auto-detects and tries to load JSON files directly, bypassing herbench.py. By declaring config names without data_files, HF will use the herbench.py loading script instead of auto-loading JSON. This matches how other datasets with custom loading scripts work on HF. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>a6c22d4 -
Fix Dataset Viewer schema mismatch (v1.0.3)
Remove non-existent fields from schema that were causing FeaturesError: - Removed evidence_count field (not in actual JSON data) - Removed difficulty field (not in actual JSON data) Schema now matches actual data structure: - Core fields: question_id, video_id, video_path, question, choices, answer, answer_index, answer_text, task_type - Metadata fields: source_dataset, duration, resolution - Full metadata preserved in metadata_json as JSON string This fixes the "Failed to convert pandas DataFrame to Arrow Table" error in the HF Dataset Viewer. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>09f2789 -
Fix Dataset Viewer and clean up repository
- Update herbench.py to v1.0.2 with robust schema handling - Fix metadata field inconsistencies across task types - Add proper type normalization for all fields - Improve streaming compatibility for Dataset Viewer - Store full metadata as JSON string in metadata_json field - Update README.md with Dataset Viewer information - Add note about herbench.py loading script - Document stable schema handling - Update .gitignore - Add DEVELOPMENT.md to ignore list (local dev docs only) - Keep scripts/ ignored (not needed in HF dataset card) This commit ensures the HF Dataset Viewer works properly and keeps the repository clean and professional for the public dataset card. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>89563c9
-
Fix HF Dataset Viewer: add hub loader script
38f2cd8 -
Upload data/video_metadata.json with huggingface_hub
Verified signature5b83e71 -
Upload data/task_metadata.json with huggingface_hub
Verified signature63595ba -
Upload data/herbench_annotations_lite.json with huggingface_hub
Verified signaturee09548a -
Upload data/herbench_annotations.json with huggingface_hub
Verified signaturec759831 -
Upload assets/Teaser_plot.jpg with huggingface_hub
Verified signaturec4937ba -
Upload assets/HERBench_logo.png with huggingface_hub
Verified signatured69ae9e -
Allow HF CLI uploads for assets/data/videos (only ignore sub-readmes/raw)
82e852d -
Upload data/video_metadata.json with huggingface_hub
Verified signature1f5e8ce -
Upload data/task_metadata.json with huggingface_hub
Verified signature911d153 -
Upload data/herbench_annotations_lite.json with huggingface_hub
Verified signature5d46c04 -
Upload data/herbench_annotations.json with huggingface_hub
Verified signature61eaea9 -
Upload assets/Teaser_plot.jpg with huggingface_hub
Verified signature63da866 -
Upload assets/HERBench_logo.png with huggingface_hub
Verified signature589850e -
Upload assets/HERBench_logo.png with huggingface_hub
Verified signature628dda9 -
Upload assets/HERBench_logo.png with huggingface_hub
Verified signature9d88728 -
Upload assets/HERBench_logo.png with huggingface_hub
Verified signature503f202 -
Upload assets/bench_stats.jpg with huggingface_hub
Verified signature7e4d551 -
Upload data/video_metadata.json with huggingface_hub
Verified signaturebd44eca -
Upload data/task_metadata.json with huggingface_hub
Verified signature0592075 -
Upload data/herbench_annotations_lite.json with huggingface_hub
Verified signature020e76b -
Upload data/herbench_annotations.json with huggingface_hub
Verified signature49273ee -
Upload assets/Teaser_plot.jpg with huggingface_hub
Verified signature109a4be -
Upload assets/HERBench_logo.png with huggingface_hub
Verified signaturec22d704 -
Upload assets/bench_stats.jpg with huggingface_hub
Verified signature019e9a5 -
Initial commit: HERBench dataset core files
9e54dc1