基于多类最大散度差的人脸表示方法.pdfVIP

32 3 Vol. 32, No. 3 2006 5 ACTA AUTOMATICA SINICA May, 2006 1) 1, 2 3 ℄2 4 1( 518055) 2 ( 230031) 3 ( 210094) 4 () (E-mail: fx song@) Æ Æ℄ Æ ORL Yale FERET ℄ Fisherfaces Eigenfaces  Fisher TP39 Face Representation Based on the Multiple-class Maximum Scatter Difference SONG Feng-Xi1,2 YANG Jing-Yu3 LIU Shu-Hai2 ZHANG David4 1(Shenzhen Graduate School, Harbin Institute of Technology, Shenzhen 518055) 2 (Artillery Academy, Hefei 230031) 3 (Nanjing University of Science and Technology, Nanjing 210094) 4 (Hong Kong Polytechnic University) (E-mail: fx song@) Abstract In this paper we extend the maximum scatter difference discriminant criterion which is proposed for binary classification to the multiple-class maximum scatter difference discriminant criterion. Based on this new criterion we establish a novel face representation method. The facial feature extraction method based on the multiple-class maximum scat- ter difference discriminant criterion effectively avoids the small sample size problem which always brings troubles to conventional discriminant analysis methods when they are ap- plied to face recognition tasks. Experimental results conducted on international benchmark datasets such as ORL, Yale, and FERET face image databases demonstrate that the novel face representation method is promising i