Incremental learning by physical coaching

Contributor: Dongheui Lee 

In order to achieve intuitive teaching of natural motions, we proposed a method for incremental learning by using physical interaction and implemented and evaluated using the DLR’s humanoid upper-body robot Justin shown in tha above Figure. In order to ensure synchronization of complex whole body motions on a humanoid robot, our imitation learning procedure starts with observation learning prior to kinesthetic motion refinements. A customized impedance controller allows to combine tracking of motion primitives in free-space with a kinesthetic modification by a human supervisor which avoids accidental disturbance. 

 

 


Related Publications

Journal articles

  1. Dongheui Lee and Christian Ott, Incremental Kinesthetic Teaching of Motion Primitives Using the Motion Refinement Tube. Autonomous Robots, 31(2):115–131, 2011

Peer-reviewed conference papers

  1. Christian Ott, Bernd Henze, and Dongheui Lee, Kinesthetic teaching of humanoid motion based on whole-body compliance control with interaction-aware balancing, in Proc. IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pp. 4615-4621, 2013
  2. Dongheui Lee and Christian Ott, Iterative Motion Primitive Learning and Refinement by Complaint Motion Control, in IEEE International Conference on Humanoid Robots (HUMANOIDS) workshop Humanoid Robots Learning from Human Interaction, USA, 2010
  3. Dongheui Lee and Christian Ott, Incremental Motion Primitive Learning by Physical Coaching Using Impedance Control, in Proc. IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pp. 4133-4140, Taipei, Taiwan, Oct. 2010
  4. Dongheui Lee, Christian Ott, Yoshihiko Nakamura and Gerd Hirzinger, Physical Human Robot Interaction in Imitation Learning, in Proc. IEEE International Conference on Robotics and Automation (ICRA), pp 3439-3440, 2011