GAN for Vision, KG for Relation: a Two-stage Deep Network for Zero-shot Action Recognition.
Bin SunDehui KongShaofan WangJinghua LiBaocai YinXiaonan LuoPublished in: CoRR (2021)
Keyphrases
- action recognition
- computer vision
- bag of words
- human actions
- spatial temporal
- activity recognition
- human detection
- static images
- recognition of human actions
- bag of features
- body parts
- human pose
- recognizing human actions
- view invariant action recognition
- recognizing actions
- action classification
- action detection
- depth cameras
- spatio temporal
- independent subspace analysis
- image processing
- action primitives
- video dataset
- action recognition in videos