Title :
A improved method of face recognition by kernel maximum margin criterion
Author :
Yong-zhi Li ; Jing-Yu Yang
Author_Institution :
Nanjing Univ. of Sci. & Technol., Nanjing
Abstract :
A improved method of feature extraction based on kernel maximum margin criterion (KMMC) is presented for face recognition in this paper, i.e. a simple algorithm of uncorrected optimal discriminant vectors in kernel feature space is proposed for nonlinear feature extraction. The proposed method has more powerful capability to eliminate the statistical correlation between feature vectors and its mathematical formulation is simple. Experimental results on ORL face database and YALE face database show that the new method is better than KMMC and kernel principal component analysis (KPCA) in terms of recognition rate.
Keywords :
face recognition; feature extraction; principal component analysis; face recognition; kernel feature space; kernel maximum margin criterion; kernel principal component analysis; nonlinear feature extraction; optimal discriminant vectors; statistical correlation; Computer science; Face recognition; Feature extraction; Forestry; Information science; Kernel; Principal component analysis; Scattering; Spatial databases; Support vector machines;
Conference_Titel :
Natural Computation, 2007. ICNC 2007. Third International Conference on
Conference_Location :
Haikou
Print_ISBN :
978-0-7695-2875-5
DOI :
10.1109/ICNC.2007.58