Computational Learning Group

Research Goals

The main research focus will be on imitationlearning and movment primitives for the self-organization of motor control.A second topic is on learning oculomotor and oculo-manual control. Dynamicmovement primitives will be investigated for full arm and full body movementgeneration. We will continue to develop new learning algorithms for high-demonsionalmotor control, with a focus on cerebellum-like learning andmodular motorcontrol. On the new robot, we hope to investigate the research topics abovein the example tasks of ball juggling, ball bouncing, drumming, and locomotion.

Research Topics

Imitation Learning
Statistical Learning with Neural Networks
Biomimetic Robotics
Behavioral Experiments


Stefan Schaal, USC
(Group Leader)
Shinya Kotosaka
Hiroyuki Miyamoto
Tomohiro Shibata
Ales Ude


Links to collaborators

US site
RoboticsResearch in Japan