Push Recovery and Active Balancing for Inexpensive Humanoid Robots Using RL and DRL.
Amirhossein HosseinmemarJohn AndersonJacky BaltesMeng Cheng LauZiang WangPublished in: IEA/AIE (2020)
Keyphrases
- humanoid robot
- motion planning
- reinforcement learning
- multi modal
- human robot interaction
- biologically inspired
- low cost
- fully autonomous
- motor skills
- human robot
- motion capture
- multi agent
- real robot
- motor control
- autonomous learning
- function approximation
- motion patterns
- markov decision processes
- optimal control
- human motion
- degrees of freedom
- viewpoint
- reinforcement learning algorithms
- three dimensional
- joint space
- computer vision