Presentation 2004-12-17
Approximate Solution of Matrix Principal Component Analysis (MPCA)
Kohei INOUE, Kiichi URAHAMA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Vector data are mapped to low-dimensional space with one orthonormal matrix in ordinary principal component analysis (PCA), which has been extended to 2DPCA in which matrix data are mapped into low-dimensional space with one orthonormal matrix. This paper is addressed to MPCA in which matrix data are mapped with two orthonormal matrices. We present a direct solution scheme for obtaining an approximate solution of the reconstruction error minimization problem in MPCA instead of a previous iterative solution method. The present method is applied to image compression and recognition of face images and its high performance is revealed in compression rates, recognition speeds and classification rates.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) matrix principal component analysis / higher-order singular value decomposition / image compression, face recognition
Paper # PRMU2004-145
Date of Issue

Conference Information
Committee PRMU
Conference Date 2004/12/10(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Pattern Recognition and Media Understanding (PRMU)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Approximate Solution of Matrix Principal Component Analysis (MPCA)
Sub Title (in English)
Keyword(1) matrix principal component analysis
Keyword(2) higher-order singular value decomposition
Keyword(3) image compression, face recognition
1st Author's Name Kohei INOUE
1st Author's Affiliation Faculty of Design, Kyushu University()
2nd Author's Name Kiichi URAHAMA
2nd Author's Affiliation Faculty of Design, Kyushu University
Date 2004-12-17
Paper # PRMU2004-145
Volume (vol) vol.104
Number (no) 524
Page pp.pp.-
#Pages 4
Date of Issue