Taking a Machine’s Perspective
Friday, February 21, 2020Time
Taylor Hall 203-Auditorium
Chaz Firestone is an Assistant Professor of Psychological and Brain Sciences and Director of Perception & Mind Laboratory at Johns Hopkins University
How similar is the human mind to the sophisticated machine-learning systems that can behave like it? Biologically inspired neural network models have reached human-level benchmarks in recognizing new objects and scenes. However, DeepNets can fail in surprising, fascinating, and downright bizarre ways. In this talk Firestone will show how human subjects can predict how machines will (mis)classify images, including images described in the literature as "totally unrecognizable to human eyes." Firestone suggests that human intuition is a surprisingly reliable guide to machine perception, and explores the consequences of these results for psychology, neuroscience, computer vision, “explainable AI,” and our shared cyberpunk future.
Sponsored by the Cognitive Sciences Department.
Cognitive Science Program