Approximate geometry representations and sensory fusion

Information from the external world goes through various transformations. The learning of the original neighbourhood relations of the world using only the transformed information is examined in detail. An approximate representation consists of a nite number of discretizing points and connections between neighbouring points. The goal here is to develop such an approximate representation in a self-organizing fashion. Such a system may be considered as a generalization of the Kohonen topographical map that we now equip with self-organizing neighbouring connections. For illustrative purposes an example is presented for sensory fusion: the geometry of the 3D world is learned using the outputs of two cameras.