Abstract
The Where's Waldo problem concerns how individuals can rapidly detect an object in a cluttered scene. How does the brain locate a desired object while scanning a cluttered scene? In particular, how do the brain mechanisms that govern spatially-invariant object learning and recognition also allow fast detection of objects at specific locations in a cluttered scene? A neural model provides a mechanistic explanation of how spatial and object attention, eye movement search, and invariant object learning and recognition are coordinated to solve the Where's Waldo task. This model builds on the recent ARTSCAN model of how invariant object categories are learned during eye movement search (Fazl, Grossberg, & Mingolla, 2008, Cognitive Psychology), which also simulated reaction time data showing an object advantage during spatial attention shifts (Egly, Driver, & Rafal, 1994; JEP: General; Brown & Denney, 2007, Perception & Psychophysics). The current work clarifies how the mechanisms that lead to learning of spatially-invariant categories in What stream cortical areas, such as anterior inferotemporal cortex, can link to representations of their positions in Where stream cortical areas, such as posterior parietal cortex. Thus, when an invariant object category is activated top-down by a cognitive plan, it can selectively activate the locations of sought-after object exemplars in a cluttered scene and shift spatial attention to rapidly identify them. This proposal shows how the Where's Waldo problem exploits the brain's solution of how to overcome the complementary deficiencies of What and Where stream processes by using inter-stream interactions that allow both invariant object recognition and spatially selective attention and action to occur.
Supported in part by the National Science Foundation (SBE-0354378).