Boosting Few-shot Action Recognition with Graph-guided Hybrid Matching.
Jiazheng XingMengmeng WangYudi RuanBofan ChenYaowei GuoBoyu MuGuang DaiJingdong WangYong LiuPublished in: ICCV (2023)
Keyphrases
- action recognition
- human actions
- bag of words
- activity recognition
- spatial temporal
- human detection
- body parts
- computer vision
- recognition of human actions
- action classification
- recognizing human actions
- view invariant
- graph structure
- human activities
- depth sensors
- static images
- mid level
- bag of features
- video sequences
- action primitives
- low level
- video data
- action detection
- image matching
- action recognition in videos
- video dataset
- video indexing
- key frames
- feature selection