Perceptual organization, attention and object recognition: Closing the loop

Gestalt psychologists have long argued that the whole of an object is perceived before its individual parts. This is in contrast to most theories of attention which are feature based. In this paper we present a novel architecture for object based attention, search and recognition. The algorithm consists of neurally plausible computation mechanisms and through the concepts of border-ownership and grouping neurons, provides a single mechanism to compute both bottom up and top down attention.