PGVT: Pose-Guided Video Transformer for Fine-Grained Action Recognition.
Haosong ZhangMei Chee LeongLiyuan LiWeisi LinPublished in: WACV (2024)
Keyphrases
- fine grained
- action recognition
- action detection
- human actions
- human pose
- action classification
- recognizing actions
- video dataset
- spatial temporal
- coarse grained
- recognizing human actions
- motion features
- static images
- bag of words
- human detection
- human activities
- computer vision
- recognition of human actions
- activity recognition
- body parts
- space time interest points
- bag of features
- access control
- video data
- pose estimation
- spatio temporal
- video sequences
- atomic actions
- video frames
- view invariant
- multimedia
- video content
- d objects
- keywords
- video surveillance
- probabilistic model