no code implementations • 10 Jun 2014 • Ashesh Jain, Debarghya Das, Jayesh K. Gupta, Ashutosh Saxena
We represent trajectory preferences using a cost function that the robot learns and uses it to generate good trajectories in new environments.
