Two-stream Graph Attention Convolutional for Video Action Recognition.
Deyuan ZhangHongwei GaoHailong DaiXiangbin ShiPublished in: BigDataSE (2021)
Keyphrases
- action recognition
- human actions
- action classification
- spatial temporal
- video dataset
- action detection
- recognizing human actions
- static images
- human activities
- motion features
- recognition of human actions
- activity recognition
- space time interest points
- computer vision
- human detection
- mid level
- bag of words
- video data
- multimedia
- body parts
- recognizing actions
- human pose
- view invariant
- video sequences
- bag of features
- video streams
- motion capture data
- visual features
- video shots
- graph structure
- temporal structure
- object detection
- space time
- atomic actions
- object recognition
- depth sensors
- motion history images
- video clips
- video frames