Attentional mechanisms in colour-defined motion perception. (360G-Wellcome-080205_Z_06_Z)

In visual environments, motion direction and colour of parts of an object are usually correlated. Human motion perception does not seem to automatically benefit from this correlation: motion discrimination performance in a random dot kinematogram (RDK) where all signal dots are associated with a particular colour and all noise dots are associated with another colour is not better than if no association between motion direction and colour exists. If observers have prior knowledge of the signal co lour and attend to it, motion discrimination performance increases. Two mechanisms have been proposed to explain this facilitation: a feature-based mechanism, wherein attention increases the gain of the chromatic motion signal, or, alternatively, a surface-based mechanism, wherein attention performs colour-based surface selection hence restricting the motion discrimination task to a specific colour plane. The purpose of this project is (1) to validate a novel paradigm that allows us to discrimi nate between these two mechanisms and (2) to establish their corresponding neural markers. We will exploit the fact that, for a large range of parameters, short-wavelength-sensitive (S) cones do not contribute to global motion perception, while altering the appearance of the stimulus, hence enabling surface segmentation without changing the input into the motion mechanism itself.