Self-organizing maps for visual feature representation based on natural binocular stimuli

We model the stimulus-induced development of the topography of the primary visual cortex. The analysis uses a self-organizing Kohonen model based on h igh-dimensional coding. It allows us to obtain an arbitrary number of featu re maps by defining different operators. Using natural binocular stimuli, w e concentrate on discussing the orientation, ocular dominance, and disparit y maps. We obtain orientation and ocular dominance maps that agree with ess ential aspects of biological findings. In contrast to orientation and ocula r dominance, not much is known about the cortical representation of dispari ty. As a result of numerical simulations, we predict substructures of orien tation and ocular dominance maps that correspond to disparity maps. In regi ons of constant orientation, we find a wide range of horizontal disparities to be represented. This points to geometrical relations between orientatio n, ocular dominance? and disparity maps that might be tested in experiments .