Seminar Series

A novel paradigm to study conscious visual perception

A way to study conscious perception per se is to expose the visual system to a stimulus that generates two percepts: it creates the rare opportunity of having two states in neural processing that are related to the percepts rather than to the stimulus. Here we have developed both a novel experimental paradigm and a model to study conscious 3D perception. The paradigm exploits a visual stimulus in which we manipulated two signals that are used by the brain to retrieve the 3D layout of a scene. Those signals are: binocular disparity (spatial differences between the two eyes' images) and monocular perspective. In our paradigm observers view images in which both disparity and perspective specify different orientations of a grid in 3D space. For specific combinations of disparity and perspective observers are able to select either a disparity- or a perspective-dominated percept and they are able to flip at will between the two percepts in a well-controlled way. Our paradigm is therefore a powerful tool to study conscious vision a) psychophysically and b) physiologically by means of neurophysiology and fMRI. We also found that the perceived slants are well-described by a neural-population-code model in which the percept is determined by the combined activity of three neural populations (representing disparity, perspective and residual slant signals).