May 3rd, 2012 @ 11:30pm Daniel Lee (UPenn)

Probabilistic Representations: Robotic and Neural

In order to build successful robotic systems, the proper internal representations for sensing, decision making, planning, and motor control need to be determined. I will present some representative examples of how different probabilistic state representations are used in robots for locomotion, manipulation, driving, and playing games. I will also present new theoretical work on optimal neural representations, extending recent results from Eero Simoncelli's group.

seminars/seminaritems/2012-05-03.txt · Last modified: 2012/04/26 12:14 by silberman