Robot Skill Learning from Demonstrations in Cluttered Environments

In this paper, we present importance weighted skill learning from demonstrations. Unlike prior learning from demonstration approaches, which assume obstacle-free demonstration environments, our approach is capable of learning generalizable robot skills from demonstrations provided in a cluttered environment. To further allow skill refinement as more demonstrations are provided, we present an incremental weighted skill learning approach. Experimental validation on a robot is provided.

[1]  Aude Billard,et al.  Learning Stable Nonlinear Dynamical Systems With Gaussian Mixture Models , 2011, IEEE Transactions on Robotics.

[2]  Byron Boots,et al.  Continuous-time Gaussian process motion planning via probabilistic inference , 2017, Int. J. Robotics Res..

[3]  Jun Nakanishi,et al.  Control, Planning, Learning, and Imitation with Dynamic Movement Primitives , 2003 .

[4]  Byron Boots,et al.  Towards Robust Skill Generalization: Unifying Learning from Demonstration and Motion Planning , 2017, CoRL.

[5]  Philip Chan,et al.  Toward accurate dynamic time warping in linear time and space , 2007, Intell. Data Anal..

[6]  Jan Peters,et al.  Probabilistic Movement Primitives , 2013, NIPS.

[7]  Brett Browning,et al.  A survey of robot learning from demonstration , 2009, Robotics Auton. Syst..

[8]  Aude Billard,et al.  On Learning, Representing, and Generalizing a Task in a Humanoid Robot , 2007, IEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics).

[9]  Gregory D. Hager,et al.  An incremental approach to learning generalizable robot tasks from human demonstration , 2015, 2015 IEEE International Conference on Robotics and Automation (ICRA).

[10]  Stefan Schaal,et al.  Learning coupling terms for obstacle avoidance , 2014, 2014 IEEE-RAS International Conference on Humanoid Robots.