MotionAtlas-Bench v1

MotionAtlas-Bench v1 MotionAtlas-Bench v1 is a video multiple-choice benchmark for motion and target-entity understanding. This public release contains MCQ records, answer keys, media, and target-object masks needed to reproduce the visual grounding settings. Resources Paper: Project page: Code: MotionAtlas-Data:… See the full description on the dataset page:

Source: Hugging Face Hub (maxLWSv2/motionatlas-bench). Metadata imported from the dataset’s Hub tags.