A deep reinforcement learning approach for dynamically stable inverse kinematics of humanoid robots.
S. Phani TejaParijat DewanganPooja GuhanAbhishek SarkarK. Madhava KrishnaPublished in: ROBIO (2017)
Keyphrases
- inverse kinematics
- motion planning
- humanoid robot
- reinforcement learning
- robot arm
- real robot
- end effector
- degrees of freedom
- multi modal
- path planning
- human robot interaction
- robot manipulators
- mobile robot
- state space
- neural network
- position and orientation
- human motion
- learning algorithm
- optimal control
- configuration space
- machine learning
- joint angles
- motor skills
- control law
- motion capture
- action selection
- human body
- computer vision
- real time
- walking speed