A graph convolutional neural network model with Fisher vector encoding and channel-wise spatial-temporal aggregation for skeleton-based action recognition.
Jun TangYanjiang WangSichao FuBaodi LiuWeifeng LiuPublished in: IET Image Process. (2022)
Keyphrases
- action recognition
- neural network model
- temporal aggregation
- spatial temporal
- spatio temporal
- neural network
- human actions
- artificial neural networks
- bag of words
- temporal databases
- activity recognition
- computer vision
- spatial information
- action classification
- recognizing human actions
- recognizing actions
- graph structure
- view invariant
- recognition of human actions
- spatial data
- feature vectors
- database
- action detection
- action recognition in videos
- aggregation functions
- temporal data
- pairwise
- data streams
- machine learning
- databases