Self-supervised Spatio-temporal Representation Learning for Videos by Predicting Motion and Appearance Statistics.
Jiangliu WangJianbo JiaoLinchao BaoShengfeng HeYunhui LiuWei LiuPublished in: CoRR (2019)
Keyphrases
- spatio temporal
- spatial and temporal
- dynamic textures
- learning algorithm
- recognition of human actions
- reinforcement learning
- motion patterns
- video representation
- static images
- temporal domain
- temporal segmentation
- appearance model
- perceptual information
- video sequences
- appearance features
- view invariant
- human actions
- video analysis
- motion analysis
- video frames
- object tracking
- motion estimation
- mobile robot
- image sequences