Efficient Robot Skill Learning:  Grounded Simulation Learning and Imitation Learning from Observation

Abstract

For autonomous robots to operate in the open, dynamically changing world, they will need to be able to learn a robust set of skills from relatively little experience.  This talk begins by introducing Grounded Simulation Learning as a way to bridge the so-called reality gap between simulators and the real worl in order to enable transfer learning from simulation to a real robot (sim-to-real).  It then introduces two new algorithms for imitation learning from observation that enable a robot to mimic demonstrated skills from state-only trajectories, without any knowledge of the actions selected by the Demonstrator. Grounded Simulation Learning has led to the fastest known stable walk on a widely used humanoid robot, and imitation learning from observation opens the possibility of robots learning from the vast trove of videos available online.

Notes

• The talk covers material in the following published papers on Grounded Simulation Learning [1,2,3] and Imitation Learning from Observation [4,5,6].

• Peter Stone is a Professor at the Department of Computer Science, University of Texas at Austin, and an Executive Director at Sony AI America. His personal website can be found here.