Motion capture and reinforcement learning of dynamically stable humanoid movement primitives.
Rok VugaMatjaz OgrincAndrej GamsTadej PetricNorikazu SugimotoAles UdeJun MorimotoPublished in: ICRA (2013)
Keyphrases
- motion capture
- reinforcement learning
- body movements
- human movement
- human body
- human motion
- humanoid robot
- human computer interaction
- human figure
- computer graphics
- computer animation
- multi view
- human motion capture
- pose tracking
- markerless
- state space
- joint angles
- monocular video sequences
- machine learning
- low level
- learning process
- motion capture data
- body parts
- non rigid structure from motion
- similarity measure
- three dimensional
- artificial intelligence
- human motion capture data
- virtual reality
- optimal policy
- video sequences
- image sequences
- high level
- learning algorithm