Closed-loop Perception in Humans, Rodents and Robots

Abstract: 

I’ll challenge the text-book description of perception as a sensory process, which starts at the receptors and ends somewhere in the brain. I will present the empirical data establishing the motor-sensory basis of mammalian perception, and the inescapable conclusion that perception emerges from motor-sensory-motor closed loops. I’ll describe our specific closed-loop-perception hypothesis and its predictions regarding the dynamics of perception. I will then present evidence for closed-loop low-level vision. If time permits I will show how robots can see via closed-loop dynamics.

Sponsored by the NYU-ECNU Institute of Brain and Cognitive Science at NYU Shanghai