Semi-Supervised Image-to-Video Adaptation for Video Action Recognition.
Jianguang ZhangYahong HanJinhui TangQinghua HuJianmin JiangPublished in: IEEE Trans. Cybern. (2017)
Keyphrases
- action recognition
- static images
- human actions
- action classification
- spatial temporal
- video dataset
- mid level
- semi supervised
- action detection
- video images
- motion features
- bag of features
- video frames
- recognizing human actions
- video sequences
- bag of words
- video content
- activity recognition
- input image
- video data
- human activities
- computer vision
- image content
- visual cues
- recognition of human actions
- human detection
- image retrieval
- space time interest points
- image features
- space time
- low level descriptors
- pairwise
- video clips
- image classification
- global features
- key frames
- motion history images
- high resolution
- feature points
- spatio temporal
- video scene
- human pose
- image representation
- video shots
- keypoints