SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders.
Sheng-Wei LiZi-Xiang WeiWei-Jie ChenYi-Hsin YuChih-Yuan YangJane Yung-jen HsuPublished in: CoRR (2024)
Keyphrases
- action recognition
- bag of words
- activity recognition
- human actions
- denoising
- human detection
- spatial temporal
- action classification
- recognizing human actions
- body parts
- image segmentation
- recognition of human actions
- recognizing actions
- computer vision
- view invariant
- depth sensors
- depth cameras
- bag of features
- human computer interaction
- static images
- action primitives
- d objects
- spatio temporal
- three dimensional
- independent subspace analysis