QBI Seminar: Probabilistic models of sensorimotor control and decision making

Event Details

Date:

Friday, 01 August 2014 - Friday, 01 August 2014

Time:

12:00 pm - 1:00 pm

Room:

Auditorium, Level 7

UQ Location:

Queensland Brain Institute (St Lucia)

URL:

http://learning.eng.cam.ac.uk/Public/Wolpert/Wolpert

Event category(s):

Full Description:: Speaker: Professor Daniel Wolpert
Computational and Biological Learning Lab, The University of Cambridge

Title: Probabilistic models of sensorimotor control and decision making

Abstract:

[There will be under 50% overlap with the previous day’s ICON talk so at least half will be new material]

The effortless ease with which humans move our arms, our eyes, even our lips when we speak masks the true complexity of the control processes involved. This is evident when we try to build machines to perform human control tasks. While computers can now beat grandmasters at chess, no computer can yet control a robot to manipulate a chess piece with the dexterity of a six-year-old child. I will review our work on how the humans learn to make skilled movements covering probabilistic models of learning, including cognitive tomography techniques to extract compile priors from simple measurements. I will also review our recent work showing the intimate interactions between decision making and sensorimotor control processes including active sensing through eye movements for visual texture categorisation.