ActionHub: A Large-scale Action Video Description Dataset for Zero-shot Action Recognition.
Jiaming ZhouJunwei LiangKun-Yu LinJinrui YangWei-Shi ZhengPublished in: CoRR (2024)
Keyphrases
- human actions
- action recognition
- space time interest points
- action classification
- recognizing human actions
- spatial temporal
- activity recognition
- action detection
- video dataset
- recognition of human actions
- recognizing actions
- human detection
- view invariant
- bag of words
- spatio temporal
- video sequences
- spatio temporal interest points
- body parts
- action recognition in videos
- human activities
- computer vision
- high level
- human pose
- motion features
- space time
- mid level
- motion recognition
- bag of features
- depth sensors
- motion capture data
- action primitives
- video data
- ucf sports
- low level
- human activity recognition
- video images
- visual words
- object detection
- multimedia