Colour appearance in a dynamic world. (360G-Wellcome-094595_Z_10_Z)

We argue that time-varying signals play a fundamental part in colour constancy, redefining the problem, both in terms of the information available and the nature of the neural mechanisms that might extract this information. This project considers two classes of dynamic information: (a) changes in the spectral and geometric properties of the illumination and (b) saccadic eye-movements that direct to the retina a sequence of discrete samples. Our first goal is to test observers sensitivity to tem poral cues to disambiguate properties of the illumination and of the illuminated objects. Our second goal is to consider how eye-movements modify the computational task of colour constancy and their effect on perception. Multiple mechanisms, ranging from distal to central, automatic to volitional and passive to active, contribute to maintaining constant perception of surface colour. By tracking the on-line updating of observers perceptual hypotheses we aim to (i) dissociate processes operating on different time-scales, (ii) determine the relative utility of spatial versus temporal comparisons in retinal and spatial coordinates, and (iii) assess interactions between surface colour perception and active viewing, all of which provide important information about the neural mechanisms that maintain perceptual stability, and ultimately enhance our understanding of the adaptive brain.