TY - GEN
T1 - Rank-lifting strategy based kernel regularized discriminant analysis method for face recognition
AU - Chen, Wen Sheng
AU - Yuen, Pong Chi
AU - Xie, Xuehui
N1 - Copyright:
Copyright 2011 Elsevier B.V., All rights reserved.
PY - 2010
Y1 - 2010
N2 - To address Small Sample Size (S3) problem and nonlinear problem of face recognition, this paper proposes a novel rank-lifting based kernel regularized discriminant analysis method (RL-KRDA). It first proves a rank-lifting theorem using algebraic theory. Combining a new ranklifting strategy with standby three-to-one regularization technique, the complete regularized technology is developed on the within-class scatter matrix Sw. Our regularized scheme not only adjusts the projection directions but tunes their corresponding weights as well. It is also shown that the final regularized within-class scatter matrix approaches to the original one as the regularized parameters tend to zeros. The public available database, i.e. CMU PIE face database, is selected for evaluation. Comparing with some existing kernel-based LDA methods for solving S3 problem, the proposed RL-KRDA approach gives the best performance.
AB - To address Small Sample Size (S3) problem and nonlinear problem of face recognition, this paper proposes a novel rank-lifting based kernel regularized discriminant analysis method (RL-KRDA). It first proves a rank-lifting theorem using algebraic theory. Combining a new ranklifting strategy with standby three-to-one regularization technique, the complete regularized technology is developed on the within-class scatter matrix Sw. Our regularized scheme not only adjusts the projection directions but tunes their corresponding weights as well. It is also shown that the final regularized within-class scatter matrix approaches to the original one as the regularized parameters tend to zeros. The public available database, i.e. CMU PIE face database, is selected for evaluation. Comparing with some existing kernel-based LDA methods for solving S3 problem, the proposed RL-KRDA approach gives the best performance.
KW - Face recognition
KW - Kernel method
KW - Rank lifting scheme
KW - Small sample size problem
UR - http://www.scopus.com/inward/record.url?scp=78651429902&partnerID=8YFLogxK
U2 - 10.1109/CCPR.2010.5659142
DO - 10.1109/CCPR.2010.5659142
M3 - Conference proceeding
AN - SCOPUS:78651429902
SN - 9781424472109
T3 - 2010 Chinese Conference on Pattern Recognition, CCPR 2010 - Proceedings
SP - 141
EP - 145
BT - 2010 Chinese Conference on Pattern Recognition, CCPR 2010 - Proceedings
T2 - 2010 Chinese Conference on Pattern Recognition, CCPR 2010
Y2 - 21 October 2010 through 23 October 2010
ER -