HAPACT: A benchmark for localizing physical impacts on the human body in videos
In brief
- HAPACT is a benchmark for temporal and spatial localization of physical impacts on the human body in videos.
- The dataset contains 2,800 video-query pairs and 20,000 annotated frames from movie clips.
- Results show this task remains challenging for existing models.
Researchers introduce HAPACT, a new benchmark focused on localizing physical impacts on the human body in movie clips. The dataset includes over 2,800 video-query pairs and 20,000 annotated frames, enabling evaluation of models' physical reasoning in videos.
This summary gives you the essentials. The full article, its figures and quotes are on cs.AI updates on arXiv.org.
Read the original articleAI-generated summary of the source, checked by rules. Report an error