Trends in Cognitive Sciences
Selective visual attention and perceptual coherence
Section snippets
Attention biases competition for representation in visual cortex
As stated earlier, the hierarchical structure of the visual system is characterized by two properties: increasing RF size and increasing RF complexity (Figure 1a, 3, 4, 5). When an otherwise effective sensory stimulus appears along with an otherwise ineffective sensory stimulus within a large RF, should the neuron's response be strong (reflecting the influence of the effective stimulus) or weak (reflecting the influence of the ineffective stimulus)? In other words, which of these stimuli will
Cortical computation of attentional priority
Although the biased competition account provides a useful theoretical framework in which to understand attentional modulations, it leaves open the neural mechanisms by which attentional control is implemented. Many psychological and computational models of attention posit an ‘attentional priority map’ that reflects the distribution of attention across the visual scene 25, 26, 27 (see Box 1). On some accounts, the stimulus array is initially filtered to form a bottom-up (or ‘stimulus-driven’)
Distributed attentional priority maps and perceptual coherence
The neurophysiological evidence reviewed in the previous section supports a view of attentional priority in which perceptual coherence fields are formed when the distributed representation of an attended stimulus comes to dominate activity within multiple topographically organized visual areas. Early regions like V1 provide high-acuity information about simple visual features and closely track the contents of the retinal image, intermediate levels like V4 and IT represent more complex feature
Switching attention by reconfiguring perceptual coherence fields
The neural representation of an attended stimulus is more robust than that of other competing objects at every level of the visual system. However, the studies reviewed above do not specify how the selected coherence field is reconfigured when a new target stimulus is specified by either stimulus-driven or voluntary attentional control factors. In the case of stimulus-driven control, the physical salience of a stimulus might override the current coherence field by strongly activating visually
The role of switch signals in altering perceptual coherence
How can a transient signal that carries no information about the target of an attention shift effectively establish a new coherence field as behavioral demands change over time? Two possibilities can be considered. First, the visual system might tend towards a chaotic or incoherent state when relatively unconstrained by selection demands; this incoherent state would be ‘equidistant’ from all possible coherent states, which would minimize reconfiguration time on average. On this account, the
Sources and targets of attentional deployments
In this article, following standard practice, we have drawn a distinction between the sources of attentional control (e.g. the transient switch signal) and the targets of those attentional control signals (the visual areas participating in a given perceptual coherence field including subcortical, occipital, parietal, and frontal regions). However, this dichotomy has typically been drawn along rather sweeping anatomical boundaries, with PPC and FEF (and perhaps SC and pulvinar) classified as
Conclusion
To understand visual perception, we must understand how the brain resolves competition among objects in the scene, and how the anatomically distributed bits of information belonging to each object are bound together. The studies reviewed here demonstrate that selective attention operates at each level of the visual hierarchy to resolve competition between multiple stimuli. Moreover, the ubiquity of these attention effects highlights a potentially larger role for selective attention in
Acknowledgements
Supported by NSF Graduate Research Fellowship to J.T.S. and NIDA grant R01-DA13165 to S.Y.
References (82)
Seeing, sensing, and scrutinizing
Vision Res.
(2000)How parallel is visual processing in the ventral pathway?
Trends Cogn. Sci.
(2004)- et al.
Feature-based attention increases the selectivity of population responses in primate visual cortex
Curr. Biol.
(2004) - et al.
Interacting roles of attention and visual salience in V4
Neuron
(2003) Competitive brain activity in visual attention
Curr. Opin. Neurobiol.
(1997)- et al.
Activity in posterior parietal cortex is correlated with the relative subjective desirability of action
Neuron
(2004) Climbing the cortical ladder from sensation to perception
Trends Cogn. Sci.
(2003)- et al.
View from the top: hierarchies and reverse hierarchies in the visual system
Neuron
(2002) Transient attention enhances perceptual performance and FMRI response in human visual cortex
Neuron
(2005)A saliency map in primary visual cortex
Trends Cogn. Sci.
(2002)