Topographic independent component analysis as a model of V1 organization and receptive fields

Independent component analysis (ICA) has been recently used as a model of natural image statistics and V1 simple cell receptive "elds. Here we show how to extend the ICA model to explain V1 topography as well. This is done by relaxing the independence assumption and ordering the basis vectors so that vectors with strong higher-order correlations are near each other. This is a new principle of topographic organization, and may be more relevant to natural image statistics than the more conventional topographic ordering based on Euclidean distances. For example, our ordering leads to simultaneous emergence of complex cell properties: topographic neighborhoods act like complex cells. 2001 Elsevier Science B.V. All rights reserved.