Reinforcement Learning for Bipedal Gait with MAX-E2 Humanoid Robot.
David Yanguas-RojasEduardo Mojica-NavaAlben CardenasPublished in: Int. J. Humanoid Robotics (2022)
Keyphrases
- humanoid robot
- walking speed
- reinforcement learning
- real robot
- motion capture
- motion planning
- multi modal
- biologically inspired
- learning algorithm
- function approximation
- human robot
- gait recognition
- human robot interaction
- human motion
- imitation learning
- reinforcement learning algorithms
- model free
- state space
- rough terrain
- markov decision processes
- manipulation tasks
- motor skills
- legged robots
- fully autonomous
- real time
- action space
- body movements
- multi agent
- feature extraction
- pattern generator