Multi-scale Spatiotemporal Information Fusion Network for Video Action Recognition.
Yutong CaiWeiyao LinJohn SeeMing-Ming ChengGuangcan LiuHongkai XiongPublished in: VCIP (2018)
Keyphrases
- action recognition
- information fusion
- human actions
- action classification
- spatial temporal
- multiscale
- video dataset
- action detection
- recognizing human actions
- multi source
- space time
- data fusion
- human activities
- recognition of human actions
- bag of words
- activity recognition
- static images
- motion features
- computer vision
- soft computing
- body parts
- spatial and temporal
- human detection
- spatio temporal
- video data
- space time interest points
- view invariant
- recognizing actions
- real time
- video frames
- motion history images
- computational intelligence
- action recognition in videos
- human pose
- video content
- low level
- atomic actions
- video clips
- moving objects
- video retrieval