An adaptive deep reinforcement learning framework enables curling robots with human-like performance in real-world conditions.
Dong-Ok WonKlaus-Robert MüllerSeong-Whan LeePublished in: Sci. Robotics (2020)
Keyphrases
- framework enables
- real world conditions
- reinforcement learning
- human robot interaction
- real time
- real robot
- robotic control
- robot control
- lighting conditions
- multi robot
- robotic systems
- state space
- head movements
- humanoid robot
- mobile robot
- learning algorithm
- machine learning
- user interface
- action selection
- feature space
- object recognition
- document repository