Object recognition by artificial cortical maps

Object recognition is at the same time the most valuable outcome of the human visual system, and the less understood yet; despite being vision certainly the most studied function of the brain. There is indeed a relatively good knowledge of several processes taking place in the cortical visual areas, supporting the recognition capability, like orientation discrimination and color constancy. This work proposes a model of the development of the object recognition capability, based on two main theoretical principles. First is that recognition does not imply any sort of geometrical reconstruction, it is fully driven by the two dimensional view captured by the retina. The second assumption is that all the processing functions involved in recognition are not genetically determined and hardwired in the neural circuits, but are the result of interactions between epigenetic influences and the basic neural plasticity mechanisms. The model is organized in modules roughly related with the main visual biological areas, and is implemented mainly using the LISSOM architecture, a recent neural self-organization map model simulating the effect of intercortical lateral connections.