Early Action Recognition With Category Exclusion Using Policy-Based Reinforcement Learning.
Junwu WengXudong JiangWei-Long ZhengJunsong YuanPublished in: IEEE Trans. Circuits Syst. Video Technol. (2020)
Keyphrases
- action recognition
- reinforcement learning
- optimal policy
- human actions
- bag of words
- markov decision processes
- activity recognition
- partially observable
- recognition of human actions
- body parts
- computer vision
- action classification
- state space
- human detection
- reward function
- dynamic programming
- recognizing human actions
- spatial temporal
- human activities
- recognizing actions
- learning algorithm
- machine learning
- object categories
- independent component analysis
- robust tracking
- image features
- depth sensors
- action primitives
- keywords
- feature extraction