Watching Unlabeled Video Helps Learn New Human Actions from Very Few Labeled Snapshots.
Chao-Yeh ChenKristen GraumanPublished in: CVPR (2013)
Keyphrases
- human actions
- action recognition
- spatio temporal
- video sequences
- space time
- action classification
- human activities
- human motion
- training data
- partially labeled
- supervised learning
- spatial temporal
- visual features
- activity recognition
- recognition of human actions
- recognizing human actions
- recognizing actions
- visual data
- motion features
- active learning
- semi supervised learning
- video shots
- action sequences
- view invariant
- semi supervised
- training set
- unlabeled data
- labeled data
- key frames
- motion recognition
- computer vision
- action detection
- event detection
- web videos
- video analysis
- video content
- video frames
- video data
- unsupervised learning
- image sequences