Video Action Recognition Based on Deeper Convolution Networks with Pair-Wise Frame Motion Concatenation.
Yamin HanPeng ZhangTao ZhuoWei HuangYanning ZhangPublished in: CVPR Workshops (2017)
Keyphrases
- action recognition
- recognizing human actions
- recognition of human actions
- motion features
- static images
- human actions
- pairwise
- video frames
- action classification
- human movements
- key frames
- motion history images
- spatial temporal
- mid level
- video dataset
- motion capture data
- bag of words
- action recognition in videos
- temporal coherence
- activity recognition
- action detection
- space time
- consecutive frames
- view invariant
- human detection
- video sequences
- human activities
- video data
- spatial and temporal
- action primitives
- computer vision
- body parts
- video content
- motion recognition
- space time interest points
- human motion
- video shots
- moving camera
- object motion
- bag of features
- video clips
- video streams
- motion estimation
- human pose
- video surveillance
- pose estimation
- recognizing actions
- visual features
- optical flow