2 ROLES FOR ATTENTION IN SHAPE PERCEPTION - A STRUCTURAL DESCRIPTION MODEL OF VISUAL SCRUTINY

This paper presents MetriCat, a model of the human capacity to recogni ze objects both as members of a general class (e.g. ''chair'') and as specific instances (''my office chair''), and of the role of visual at tention in this capacity. MetriCat represents the attributes of an obj ect's parts and their relations in a non-linear fashion that provides a natural basis for recognition at both the class and instance levels (Stankiewicz & Hummel, 1996). Like previous structural description mod els (e.g. Hummel & Biederman, 1992), MetriCat represents part attribut es and relations independently, dynamically binding them into structur al descriptions. The resulting representation suggests two roles for v isual attention in shape recognition: attention for binding and attent ion for signal-to-noise control. MetriCat implements both functions as special cases of a single mechanism for controlling the synchrony rel ations among units representing separate object parts. The model accou nts for the time course of class-and instance-level classification, an d makes several predictions about the relationships between attention, time and levels of classification.