Hybrid features for skeleton-based action recognition based on network fusion.
Zhangmeng ChenJunjun PanXiaosong YangHong QinPublished in: Comput. Animat. Virtual Worlds (2020)
Keyphrases
- action recognition
- static images
- spatial temporal
- human actions
- bag of words
- mid level
- bag of features
- motion features
- human detection
- low level
- body parts
- activity recognition
- space time interest points
- image features
- action classification
- feature space
- human movements
- recognition of human actions
- action primitives
- computer vision
- co occurrence
- multiple features
- deformable part models
- human pose
- spatio temporal
- global features
- feature detection
- sensor networks
- action detection
- feature vectors
- motion history images
- object recognition
- feature extraction