Demystifying the richness of visual awareness

Speaker: Ruth Rosenholtz, NVIDIA

Abstract: As humans, we experience a rich visual percept. However, we often perform poorly when probed on the details of that percept, suggesting that vision is impoverished. Philosophers and vision scientists have puzzled over this seeming contradiction. Is the rich percept a mere illusion? And if so, then how does real-world vision work as well as it does? I will argue for two important pieces of the solution. First, peripheral vision efficiently encodes its inputs using a scheme that preserves a great deal of useful information, while losing the information needed for typical tasks used to probe the details. Second, many tasks used to demonstrate impoverished vision may be inherently difficult; poor performance on these tasks may indicate general-purpose limits on task complexity. Together, these two components can make sense of a wide variety of phenomena, including vision’s marvelous successes, its quirky failures, and our rich percept of the visual world.

Biography: Ruth Rosenholtz is a Principal Research Scientist at NVIDIA. She got her Ph.D. in EECS (Computer Vision) from UC Berkeley. She was previously a member of CSAIL and MIT’s Department of Brain and Cognitive Sciences, having joined MIT in 2003 after 7 years at the Palo Alto Research Center (formerly Xerox PARC). Her work focuses on developing predictive models of human visual processing, including visual search, perceptual organization, visual clutter, peripheral vision, and attention. In addition, she has worked on applying understanding of human vision to image fidelity (NASA Ames), and to design of user interfaces and information visualizations (Xerox PARC and MIT).

Chair: Prof. Piotr Didyk