Channel-wise Topology Refinement Graph Convolution for Skeleton-Based Action Recognition.
Yuxin ChenZiqi ZhangChunfeng YuanBing LiYing DengWeiming HuPublished in: ICCV (2021)
Keyphrases
- action recognition
- topological information
- human actions
- bag of words
- human detection
- computer vision
- recognizing human actions
- action classification
- activity recognition
- recognition of human actions
- independent subspace analysis
- video dataset
- body parts
- human pose
- spatial temporal
- bag of features
- action recognition in videos
- video sequences
- static images
- pairwise
- view invariant
- action detection
- action primitives
- shape analysis
- spatial information
- d objects
- view invariant action recognition