Modeling Parallel Information Processing in Goal-Directed Vision

Alejandro Lleras, professor of psychology, faculty member in Mechanisms of Cognitive Control

Abstract:
When we open our eyes, our eyes start processing visual signals, in parallel, over the entire scene in front of us. We also have a strong phenomenology that vision is indeed occurring in parallel, that is, that we are seeing (and processing) everything in front of us simultaneously. Yet, in spite of its apparent importance, there is little systematic study of early visual parallel processing in psychology and, more generally, vision scientists have long neglected the study of the temporal dynamics of this stage of information processing (that is, the study  of how long it takes to visually process a scene). This is (we believe in part) the result of erroneous intuitions regarding parallel processing, in particular, the belief that early parallel processing is unaffected by visual complexity, such that processing a scene with few elements (or with simpler elements) ought to take the same time as one with more elements (or with more complex ones). Here I will present a line of research from our lab demonstrating that this assumption is wrong: early parallel processing is systematically affected by visual complexity in a lawful fashion. Further, we present an approach that combines computational simulations with behavioral investigations to infer what the underlying architecture of early parallel processing might be. Our approach allows us (1) to predict what processing times will be for novel scenes and (2) to identify new cognitive/visual processes that produce systematic deviations from our model’s predictions.

Alejandro Lleras is presently professor in psychology. He is also a part-time faculty member at the Beckman Institute. Prior to coming to Illinois in 2004, he was an NSF postdoctoral fellow at the University of British Columbia in Vancouver, BC.

His primary area of work is focused on understanding attention, both how it works and how it fails, mostly in the visual domain. For example, recent investigations have looked at issues such as temporal attention, why we tend to get distracted when we concentrate over long periods of time, how we tune out the world when we are trying to concentrate, and how it is that we find objects in a scene.

Since 2011, he has also served as associate editor at the Journal of Experimental Psychology:General, one of the premier experimental psychology journals. He is a member of the National Science Foundation College of Reviewers and serves in the Diversity Committee of the Psychonomic Society.