Representing Videos as Discriminative Sub-graphs for Action Recognition.
Dong LiZhaofan QiuYingwei PanTing YaoHouqiang LiTao MeiPublished in: CoRR (2022)
Keyphrases
- action recognition
- action classification
- human actions
- recognition of human actions
- recognizing actions
- video dataset
- mid level features
- ucf sports
- mid level
- bag of words
- human activities
- static images
- spatio temporal interest points
- human detection
- activity recognition
- view invariant
- motion features
- action detection
- max margin
- computer vision
- space time interest points
- recognizing human actions
- body parts
- action recognition in videos
- deformable part models
- motion history images
- spatial temporal
- human object interactions
- human pose
- feature extraction
- spatio temporal
- bag of features
- multi label
- pose estimation
- semi supervised
- video frames
- video sequences
- class specific