Dartmouth Events

Psychological and Brain Sciences Colloquium

Wolford Lecturer, Nancy Kanwisher, PhD, Massachusetts Institute of Technology

Friday, September 25, 2020
3:30pm – 4:30pm
Intended Audience(s): Alumni, Faculty, Postdoc, Staff, Students-Graduate, Students-Undergraduate
Categories: Lectures & Seminars

Please join us on Friday, September 25, 2020, at 3:30 p.m., for a virtual colloquium given by Wolford Lecturer, Nancy Kanwisher, Walter A. Rosenblith Professor of Cognitive Neuroscience in the Department of Brain and Cognitive Sciences at the Massachusetts Institute of Technology and Investigator at the McGovern Institute for Brain Research.

Title: Category Selectivity in the Ventral Visual Pathway: Computational Models and Developmental Origins

Abstract:  Despite extensive evidence for category selectivity in the ventral visual pathway, these claims of lack computational precision and remain vulnerable to empirical refutation. I will first describe work in which we constructed deep convolutional artificial neural network (ANN)-based encoding models that accurately predict the observed response to novel images in the FFA, PPA, and EBA, outperforming descriptive models and experts. We then use these models to subject claims of category selectivity to their strongest tests to date, by screening for and synthesizing images predicted to produce the highest response in each region.  We find that these high-response-predicted images were all obvious members of the hypothesized preferred category for that region, strengthening evidence for the claimed specificity of each region. I will then consider the question of how these selective regions get wired up in development. I will first show new evidence that these regions are all present, and already category selective, in young infants, and then consider the roles of domain-specific predispositions, visual experience, and long-range connectivity in their development.

Zoom details will be provided at a later date.

For more information, contact:
Michelle Powers

Events are free and open to the public unless otherwise noted.