Computational modelling of visual attention

Five important trends have emerged from recent work on computational models of focal visual attention that emphasize the bottom-up, image-based contro l of attentional deployment. First, the perceptual saliency of stimuli crit ically depends on the surrounding context. Second, a unique 'saliency map' that topographically encodes for stimulus conspicuity over the visual scene has proved to be an efficient and plausible bottom-up control strategy Thi rd, inhibition of return, the process by which the currently attended locat ion is prevented from being attended again, is a crucial element of attenti onal deployment. Fourth, attention and eye movements tightly interplay, pos ing computational challenges with respect to the coordinate system used to control attention. And last, scene understanding and object recognition str ongly constrain the selection of attended locations. Insights from these fi ve key areas provide a framework for a computational and neurobiological un derstanding of visual attention.