Dynamic video mix-up for cross-domain action recognition.
Han WuChunfeng SongShaolong YueZhenyu WangJun XiaoYanyang LiuPublished in: Neurocomputing (2022)
Keyphrases
- action recognition
- cross domain
- human actions
- action classification
- spatial temporal
- video dataset
- recognition of human actions
- action detection
- static images
- recognizing human actions
- activity recognition
- bag of words
- motion features
- human activities
- space time interest points
- computer vision
- video data
- text categorization
- video sequences
- transfer learning
- knowledge transfer
- motion history images
- action recognition in videos
- video content
- recognizing actions
- event recognition
- video frames
- space time
- view invariant
- spatio temporal
- video clips
- atomic actions