Mitigating Context Bias in Action Recognition via Skeleton-Dominated Two-Stream Network.
Qiankun LiXiaolong HuangYuwen LuoXiaoyu HuXinyu SunZengfu WangPublished in: AMC-SME (2023)
Keyphrases
- action recognition
- human actions
- activity recognition
- bag of words
- action classification
- computer vision
- human detection
- bag of features
- recognizing human actions
- recognition of human actions
- spatial temporal
- body parts
- context sensitive
- static images
- human pose
- mid level
- video dataset
- action recognition in videos
- view invariant
- human activities
- data streams
- depth sensors
- independent subspace analysis
- space time interest points
- recognizing actions
- view invariant action recognition
- action primitives
- motion history images
- action detection
- depth cameras
- human activity recognition
- stream processing
- wireless sensor networks
- viewpoint