Neuroscience 2004 Abstract
| Presentation Number: | 751.10 |
|---|---|
| Abstract Title: | Discrimination performance with realistic images is correlated with selectivity of inferior temporal (IT) neurons. |
| Authors: |
Allred, S. R.*1
; Jagadeesh, B.1,2
1Neurobiology & Behav, Univ Washington, Seattle, WA 2Physiology & Biophysics, Univ Washington, Seattle, WA |
| Primary Theme and Topics |
Sensory Systems - Vision -- Visual cortex: Objects and faces |
| Session: |
751. Visual Cortex: Processing and Seeing Objects Poster |
| Presentation Time: | Tuesday, October 26, 2004 2:00 PM-3:00 PM |
| Location: | San Diego Convention Center - Hall A-H, Board # S2 |
| Keywords: | OBJECT RECOGNITION, EXTRASTRIATE CORTEX, PERCEPTION |
Neurons in IT cortex are thought to underlie primate ability to recognize and discriminate among realistic images, but the mechanisms responsible for this process are poorly understood. We propose that IT response similarity is correlated with the discriminability of images, which reflects perceptual or psychological similarity (Allred and Jagadeesh, SFN abstracts, 2002). To test our hypothesis, we manipulated discriminability of realistic image pairs in a delayed-match-to-sample task, (Allred et. al., 2003, VSS abstracts) and measured corresponding changes in neural response similarity. Discrimination performance was manipulated by changing sample duration (SOA) from 0 to 320 ms, and by masking or not masking the sample (Kovacs et. al., 1995; Rolls et. al., 1999). We collected simultaneous behavioral and neural data, and optimized IT selectivity on each day by using two new images (effective and ineffective) and a pair-unique mask. Increasing SOA and removing the mask improved both behavioral and neural performance. We analyzed neurons that were significantly selective for the effective and ineffective image pair at the longest SOA. The mask alone (0 ms SOA) elicited a neural response that was, on average, midway between the response to effective and ineffective images at the longest SOA. As SOA increased, response to the effective image increased while response to the ineffective image decreased. This improvement in neural selectivity was comparable in magnitude to the improvement in behavioral performance. On average, neural responses on trials without the mask were more selective than on trials with the mask; behavioral performance was also higher. Changes in measured perceptual similarity (induced by manipulating SOA and removing the mask) were accompanied by comparable changes in neural response similarity.
Supported by Whitehall Foundation, McKnight Foundation, Sloan Foundation, to BJ; NEI to SRA
Sample Citation:
[Authors]. [Abstract Title]. Program No. XXX.XX. 2004 Neuroscience Meeting Planner. San Diego, CA: Society for Neuroscience, 2004. Online.
Copyright © 2004-2026 Society for Neuroscience; all rights reserved. Permission to republish any abstract or part of any abstract in any form must be obtained in writing by SfN office prior to publication.
