Real-world reinforcement learning for autonomous humanoid robot docking.
Nicolás Navarro-GuerreroCornelius WeberPascal SchroeterStefan WermterPublished in: Robotics Auton. Syst. (2012)
Keyphrases
- humanoid robot
- real robot
- reinforcement learning
- real world
- motion planning
- multi modal
- biologically inspired
- imitation learning
- human robot interaction
- fully autonomous
- autonomous learning
- model free
- robotic systems
- walking speed
- motion capture
- function approximation
- body movements
- human robot
- multi agent
- manipulation tasks
- pattern generator
- human motion
- reinforcement learning algorithms
- action selection
- partial observability
- optimal policy
- state space
- spatio temporal
- joint space
- viewpoint
- high dimensional
- computer vision