Fast Second-Order Orthogonal Tensor Subspace Analysis for Face Recognition

Tensor subspace analysis (TSA) and discriminant TSA (DTSA) are two effective two-sided projection methods for dimensionality reduction and feature extraction of face image matrices. However, they have two serious drawbacks. Firstly, TSA and DTSA iteratively compute the left and right projection matr...

Full description

Saved in:
Bibliographic Details
Main Authors: Yujian Zhou, Liang Bao, Yiqin Lin
Format: Article
Language:English
Published: Wiley 2014-01-01
Series:Journal of Applied Mathematics
Online Access:http://dx.doi.org/10.1155/2014/871565
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1832556559847129088
author Yujian Zhou
Liang Bao
Yiqin Lin
author_facet Yujian Zhou
Liang Bao
Yiqin Lin
author_sort Yujian Zhou
collection DOAJ
description Tensor subspace analysis (TSA) and discriminant TSA (DTSA) are two effective two-sided projection methods for dimensionality reduction and feature extraction of face image matrices. However, they have two serious drawbacks. Firstly, TSA and DTSA iteratively compute the left and right projection matrices. At each iteration, two generalized eigenvalue problems are required to solve, which makes them inapplicable for high dimensional image data. Secondly, the metric structure of the facial image space cannot be preserved since the left and right projection matrices are not usually orthonormal. In this paper, we propose the orthogonal TSA (OTSA) and orthogonal DTSA (ODTSA). In contrast to TSA and DTSA, two trace ratio optimization problems are required to be solved at each iteration. Thus, OTSA and ODTSA have much less computational cost than their nonorthogonal counterparts since the trace ratio optimization problem can be solved by the inexpensive Newton-Lanczos method. Experimental results show that the proposed methods achieve much higher recognition accuracy and have much lower training cost.
format Article
id doaj-art-8a1f287526de417884450955c2767726
institution Kabale University
issn 1110-757X
1687-0042
language English
publishDate 2014-01-01
publisher Wiley
record_format Article
series Journal of Applied Mathematics
spelling doaj-art-8a1f287526de417884450955c27677262025-02-03T05:44:58ZengWileyJournal of Applied Mathematics1110-757X1687-00422014-01-01201410.1155/2014/871565871565Fast Second-Order Orthogonal Tensor Subspace Analysis for Face RecognitionYujian Zhou0Liang Bao1Yiqin Lin2Department of Mathematics and Computational Science, Institute of Computational Mathematics, Hunan University of Science and Engineering, Yongzhou 425100, ChinaDepartment of Mathematics, East China University of Science and Technology, Shanghai 200237, ChinaDepartment of Mathematics and Computational Science, Institute of Computational Mathematics, Hunan University of Science and Engineering, Yongzhou 425100, ChinaTensor subspace analysis (TSA) and discriminant TSA (DTSA) are two effective two-sided projection methods for dimensionality reduction and feature extraction of face image matrices. However, they have two serious drawbacks. Firstly, TSA and DTSA iteratively compute the left and right projection matrices. At each iteration, two generalized eigenvalue problems are required to solve, which makes them inapplicable for high dimensional image data. Secondly, the metric structure of the facial image space cannot be preserved since the left and right projection matrices are not usually orthonormal. In this paper, we propose the orthogonal TSA (OTSA) and orthogonal DTSA (ODTSA). In contrast to TSA and DTSA, two trace ratio optimization problems are required to be solved at each iteration. Thus, OTSA and ODTSA have much less computational cost than their nonorthogonal counterparts since the trace ratio optimization problem can be solved by the inexpensive Newton-Lanczos method. Experimental results show that the proposed methods achieve much higher recognition accuracy and have much lower training cost.http://dx.doi.org/10.1155/2014/871565
spellingShingle Yujian Zhou
Liang Bao
Yiqin Lin
Fast Second-Order Orthogonal Tensor Subspace Analysis for Face Recognition
Journal of Applied Mathematics
title Fast Second-Order Orthogonal Tensor Subspace Analysis for Face Recognition
title_full Fast Second-Order Orthogonal Tensor Subspace Analysis for Face Recognition
title_fullStr Fast Second-Order Orthogonal Tensor Subspace Analysis for Face Recognition
title_full_unstemmed Fast Second-Order Orthogonal Tensor Subspace Analysis for Face Recognition
title_short Fast Second-Order Orthogonal Tensor Subspace Analysis for Face Recognition
title_sort fast second order orthogonal tensor subspace analysis for face recognition
url http://dx.doi.org/10.1155/2014/871565
work_keys_str_mv AT yujianzhou fastsecondorderorthogonaltensorsubspaceanalysisforfacerecognition
AT liangbao fastsecondorderorthogonaltensorsubspaceanalysisforfacerecognition
AT yiqinlin fastsecondorderorthogonaltensorsubspaceanalysisforfacerecognition