ViA: View-invariant Skeleton Action Representation Learning via Motion Retargeting.
Di YangYaohui WangAntitza DantchevaLorenzo GarattoniGianpiero FrancescaFrançois BrémondPublished in: CoRR (2022)
Keyphrases
- view invariant
- action recognition
- human motion analysis
- learning process
- multi view
- human actions
- single view
- cross view action recognition
- recognizing human actions
- spatio temporal
- learning algorithm
- space time
- motion capture
- human motion
- object recognition
- hand gestures
- low level
- active learning
- optical flow
- reinforcement learning
- image sequences