BIOLOGICALLY MOTIVATED COMPUTATIONALLY INTENSIVE APPROACHES TO IMAGE PATTERN-RECOGNITION

This paper presents some of the research activities of the research gr oup in vision as a grand challenge problem whose solution is estimated to need the power of Tflop/s computers and for which computational me thods have yet to be developed. The concerned approaches are biologica lly motivated, in that we try to mimic and use mechanisms employed by natural vision systems, more specifically the visual system of primate s. Visual information representations which are motivated by the funct ion of the primary visual cortex, more specifically by the function of so-called simple cells, are computed. Three different methods for usi ng such representations to solve image pattern recognition problems ar e presented. These are: (i) extraction and comparison of lower-dimensi on representations, (ii) computing optimal mappings of an image onto o ther images by optic flow techniques and (iii) application of a seif-o rganising neural network classifier. The problems of automatic recogni tion and classification of visual patterns, in particular the discrimi nation of human faces, are used to test the usefulness and feasibility of these approaches.