Can robots learn from humans?
Imitation seems to be one of the intrinsic motivation mechanisms for humans. Through so-called imitation learning process, an expert can easily demonstrate skills in a natural way and a student can learn the general concept of the skills by observing the teacher. In order to develop intelligent robotic assistants for humans, we address autonomous and incremental motor skill learning and control of robots. Imitation learning in robotics turns out to be an efficient way to teach robots new skills through human guidance, which can reduce time and cost to program the robot.
In the talk, incremental learning in different problem domains will be discussed, including the refinement of learned skills via heterogeneous learning modalities, enhancement of human-robot cooperation tasks over time, and improvement of stability by iterative learning control in bipedal walking. The proposed algorithms are verified via implementations on a human skeleton model and multiple robotic systems including full size humanoid robots.
Dieser Vortrag ist auf Englisch.