作者: Thomas D. Arbuckle
DOI:
关键词:
摘要: A system comprising a neural network, or computer, implementing feature detection and statistical procedure, together with fuzzy logic for solving the problem of recognition faces other objects) at multiple resolutions is described. plurality previously described systems recognizing (or which use local autocorrelation coefficients linear discriminant analysis are trained on data set to recognize facial images each particular resolution. In second training stage, tested in presented have matching resolution those first set, performance this stage being used train combination technique, that integrals. Finally, test results from classifiers combined using produce an aggregated whose higher than any individual shows very good relative all known face recognitior operate similar types testing data, system, however, not limited able be applied objects.