Motion Planner Augmented Reinforcement Learning for Robot Manipulation in Obstructed Environments.
Jun YamadaYoungwoon LeeGautam SalhotraKarl PertschMax PfluegerGaurav S. SukhatmeJoseph J. LimPeter EnglertPublished in: CoRR (2020)
Keyphrases
- reinforcement learning
- humanoid robot
- motion control
- manipulation tasks
- mobile robot
- autonomous robots
- motion planning
- robot control
- real robot
- robotic systems
- end effector
- robotic tasks
- autonomous navigation
- parallel robot
- inverse kinematics
- unstructured environments
- robot motion
- robot behavior
- position and orientation
- human robot interaction
- action selection mechanism
- configuration space
- monocular vision
- service robots
- camera motion
- motion model
- path planner
- dynamic environments
- robot navigation
- motion analysis
- space time
- control signals
- image sequences
- perceptual aliasing
- sagittal plane
- action selection
- degrees of freedom
- optical flow
- vision system
- motion estimation
- state space
- autonomous learning
- real environment
- robot arm
- path planning
- function approximation
- domain independent
- robot soccer
- control system
- plan execution
- hand eye calibration
- model free
- collision free
- human motion
- heuristic search
- dynamic programming
- moving objects
- physical constraints
- robot manipulators
- machine learning
- multi agent
- computer vision
- learning algorithm