DVANet: Disentangling View and Action Features for Multi-View Action Recognition.
Nyle SiddiquiPraveen TirupatturMubarak ShahPublished in: CoRR (2023)
Keyphrases
- multi view
- view invariant
- action recognition
- multiple views
- human actions
- single view
- multiple viewpoints
- action classification
- recognizing human actions
- recognizing actions
- recognition of human actions
- d objects
- space time interest points
- action detection
- semi supervised
- depth map
- range images
- surface reconstruction
- three dimensional
- multiple cameras
- viewpoint
- activity recognition
- computer vision
- feature extraction
- motion history images
- motion capture
- optical flow
- human motion
- view synthesis
- active learning
- body parts
- feature vectors
- spatio temporal
- human pose
- object recognition
- video sequences
- human activities