Action Recognition in Untrimmed Videos with Composite Self-Attention Two-Stream Framework.
Dong CaoLisha XuHaibo ChenPublished in: CoRR (2019)
Keyphrases
- action recognition
- video dataset
- human actions
- action classification
- recognition of human actions
- view invariant
- bag of words
- activity recognition
- human detection
- recognizing actions
- computer vision
- human activities
- recognizing human actions
- video sequences
- spatial temporal
- static images
- motion history images
- space time interest points
- motion features
- video frames
- action detection
- max margin
- bag of features
- human motion
- action recognition in videos
- spatio temporal interest points
- image sequences