Datasets:
Tasks:
Visual Question Answering
Formats:
parquet
Languages:
English
Size:
100K - 1M
Tags:
video
multimodal
event-understanding
temporal-grounding
hierarchical-annotations
post-training
License:
| # Noncommercial use and source-specific terms | |
| EventLens-derived hierarchy annotations and proxy labels use | |
| [CC-BY-NC-SA-4.0](https://creativecommons.org/licenses/by-nc-sa/4.0/). | |
| Preserve attribution and share-alike requirements. This does not replace | |
| upstream third-party terms or license third-party video rights. | |
| | Source | Applicable terms | | |
| | --- | --- | | |
| | ET-Instruct-164K | [CC-BY-NC-SA-4.0 source card](https://huggingface.co/datasets/PolyU-ChenLab/ET-Instruct-164K) | | |
| | TimeLens-100K | [Custom license](licenses/TimeLens-LICENSE.txt): academic/noncommercial purposes, excludes commercial/production use, includes an EU restriction. Constituent third-party terms remain applicable. | | |
| | TimeRFT / TimeR1 | [Upstream project](https://github.com/xiaomi-research/time-r1) and each constituent dataset's terms. | | |
| | LLaVA-Video-178K | [Upstream dataset](https://huggingface.co/datasets/lmms-lab/LLaVA-Video-178K) and constituent media terms; academic/educational use. | | |
| All media retain source-specific terms. The code license is separate. | |