A Hybrid Fuzzy Semi-supervised Learning Algorithm for Face Recognition

摘要: In this paper, we develop a hybrid fuzzy semi supervised learning algorithm (HFSA) for face recognition, which is based on the segregation of distinctive regions that include outlier instances and its counterparts. First, it achieves distribution information each sample represented with membership degree, then grade incorporated into redefinition scatter matrices, as result, initial classification whole regular feature space obtained. Second, new semi-supervised clustering presented basis precise number clusters pattern centers have been previously obtained in discovery stage, applied order to perform classification, yielding final recognition. Experimental results conducted ORL XM2VTS databases demonstrate effectiveness proposed method.