Reversing Two-Stream Networks with Decoding Discrepancy Penalty for Robust Action Recognition.
Yunbo WangZhiyu YaoHongyu ZhuMingsheng LongJianmin WangPhilip S. YuPublished in: CoRR (2018)
Keyphrases
- action recognition
- view invariant
- human actions
- activity recognition
- human detection
- bag of words
- body parts
- computer vision
- action classification
- spatial temporal
- static images
- recognizing actions
- recognition of human actions
- video dataset
- action detection
- depth sensors
- recognizing human actions
- data streams
- spatio temporal
- viewpoint
- bag of features
- robust tracking
- partial occlusion
- depth cameras
- human activities
- motion vectors
- feature points
- object detection
- sensor networks
- view invariant action recognition