Compound Prototype Matching for Few-shot Action Recognition.
Lijin YangYifei HuangYoichi SatoPublished in: CoRR (2022)
Keyphrases
- action recognition
- human actions
- human detection
- activity recognition
- computer vision
- bag of words
- action classification
- spatial temporal
- recognizing human actions
- feature points
- recognizing actions
- recognition of human actions
- static images
- bag of features
- video sequences
- video data
- visual features
- mid level
- body parts
- view invariant
- video shots
- depth cameras
- action detection
- action recognition in videos
- independent subspace analysis
- video scene
- depth sensors
- image matching