Aligned Dynamic-Preserving Embedding for Zero-Shot Action Recognition.
Yi TianYu KongQiuqi RuanGaoyun AnYun FuPublished in: IEEE Trans. Circuits Syst. Video Technol. (2020)
Keyphrases
- action recognition
- human actions
- activity recognition
- bag of words
- human detection
- action classification
- computer vision
- recognition of human actions
- spatial temporal
- body parts
- recognizing human actions
- recognizing actions
- mid level
- human activities
- motion features
- video dataset
- independent subspace analysis
- static images
- human pose
- moving objects
- bag of features
- depth sensors
- view invariant
- multi view
- optical flow
- viewpoint
- video sequences