Meta Inverse Reinforcement Learning via Maximum Reward Sharing for Human Motion Analysis.
Kun LiJoel W. BurdickPublished in: CoRR (2017)
Keyphrases
- inverse reinforcement learning
- human motion analysis
- partially observable environments
- human body
- reward function
- preference elicitation
- gait analysis
- view invariant
- human motion
- reinforcement learning
- temporal difference
- markerless
- utility function
- state space
- machine learning
- three dimensional
- reinforcement learning algorithms
- markov decision process
- multiple agents
- pose estimation
- spatio temporal