            |
GRASP Seminar Series: Fall 2007
October 26, 11:00 a.m., Wu & Chen Auditorium, Levine Hall (3330 Walnut Street)
John Tsotsos
York University
"Different Binding Strategies for the Different Stages of Visual Recognition"
Abstract: Many
think attention needs an executive to allocate visual processing resources
in the brain (and in machines). Although the cortex exhibits substantial
plasticity, dynamic allocation of neurons seems outside its capability.
Suppose instead that the processing structure is fixed, but can be ‘tuned’ to
task needs. The only resource that can be allocated is time. But most
believe that vision can be accomplished in only one feed-forward pass,
that is, within a fixed amount of time.
I will present several arguments
to refute this. But how can this fixed cortical structure be used over
periods of time longer than one feed-forward pass? I forward the proposal
that by using multiple passes of the visual processing hierarchy, both
bottom-up and top-down, and using task information to tune the processing
prior to each pass, we can explain the different recognition behaviors
that human vision exhibits. To accomplish this, four different kinds
of binding processes are introduced and are tied directly to specific
recognition tasks and their time course. The computational framework
inherent in my Selective Tuning model of attention provides the necessary
substrate for these tasks and several examples will be shown.
Biography: John Tsotsos is a Professor in the Department of Computer Science and Engineering at York University. He received his Ph.D. in Computer Science from the University of Toronto. John's research area is computational vision with a current major focus being the modelling of visual attention.
GRASP
Laboratory
Site maintained by graspadm@grasp.cis.upenn.edu
Last update:12 September, 2007