DVAMN: Dual Visual Attention Matching Network for Zero-Shot Action Recognition.
Cheng QiZhiyong FengMeng XingYong SuPublished in: ICANN (5) (2021)
Keyphrases
- action recognition
- visual attention
- human actions
- eye tracking
- eye movements
- bag of words
- saliency map
- activity recognition
- human detection
- action classification
- vision system
- computer vision
- higher level
- body parts
- salient regions
- recognition of human actions
- recognizing human actions
- human computer interaction
- human activities
- object based visual attention
- motion features
- image matching
- feature points
- multiscale