A fast ℓ1-solver and its applications to robust face recognition

In this paper we apply a recently proposed Lagrange Dual Method (LDM) to design a new Sparse Representation-based Classification (LDM-SRC) algorithm for robust face recognition problem. The proposed approach improves the efficiency of the SRC algorithm significantly. The proposed algorithm has the f...

Full description

Bibliographic Details
Main Authors: Qiu, H., Chen, Xiaoming, Liu, W., Zhou, Guanglu, Wang, Y., Lai, J.
Format: Journal Article
Published: American Institute of Mathematical Sciences 2012
Online Access:http://hdl.handle.net/20.500.11937/10229
_version_ 1848746175072567296
author Qiu, H.
Chen, Xiaoming
Liu, W.
Zhou, Guanglu
Wang, Y.
Lai, J.
author_facet Qiu, H.
Chen, Xiaoming
Liu, W.
Zhou, Guanglu
Wang, Y.
Lai, J.
author_sort Qiu, H.
building Curtin Institutional Repository
collection Online Access
description In this paper we apply a recently proposed Lagrange Dual Method (LDM) to design a new Sparse Representation-based Classification (LDM-SRC) algorithm for robust face recognition problem. The proposed approach improves the efficiency of the SRC algorithm significantly. The proposed algorithm has the following advantages: (1) it employs the LDM ℓ1-solver to find solution of theℓ1-norm minimization problem, which is much faster than other state-of-the-art ℓ1-solvers, e.g. ℓ1-magic and ℓ1−ℓs . (2) The LDM ℓ1-solver utilizes a new Lagrange-dual reformulation of the original ℓ1-norm minimization problem, not only reducing the problem size when the dimension of training image data is much less than the number of training samples, but also making the dual problem become smooth and convex. Therefore it converts the non-smooth ℓ1-norm minimization problem into a sequence of smooth optimization problems. (3) The LDM-SRC algorithm can maintain good recognition accuracy whilst reducing the computational time dramatically. Experimental results are presented on some benchmark face databases.
first_indexed 2025-11-14T06:29:04Z
format Journal Article
id curtin-20.500.11937-10229
institution Curtin University Malaysia
institution_category Local University
last_indexed 2025-11-14T06:29:04Z
publishDate 2012
publisher American Institute of Mathematical Sciences
recordtype eprints
repository_type Digital Repository
spelling curtin-20.500.11937-102292017-09-13T14:48:43Z A fast ℓ1-solver and its applications to robust face recognition Qiu, H. Chen, Xiaoming Liu, W. Zhou, Guanglu Wang, Y. Lai, J. In this paper we apply a recently proposed Lagrange Dual Method (LDM) to design a new Sparse Representation-based Classification (LDM-SRC) algorithm for robust face recognition problem. The proposed approach improves the efficiency of the SRC algorithm significantly. The proposed algorithm has the following advantages: (1) it employs the LDM ℓ1-solver to find solution of theℓ1-norm minimization problem, which is much faster than other state-of-the-art ℓ1-solvers, e.g. ℓ1-magic and ℓ1−ℓs . (2) The LDM ℓ1-solver utilizes a new Lagrange-dual reformulation of the original ℓ1-norm minimization problem, not only reducing the problem size when the dimension of training image data is much less than the number of training samples, but also making the dual problem become smooth and convex. Therefore it converts the non-smooth ℓ1-norm minimization problem into a sequence of smooth optimization problems. (3) The LDM-SRC algorithm can maintain good recognition accuracy whilst reducing the computational time dramatically. Experimental results are presented on some benchmark face databases. 2012 Journal Article http://hdl.handle.net/20.500.11937/10229 10.3934/jimo.2012.8.163 American Institute of Mathematical Sciences fulltext
spellingShingle Qiu, H.
Chen, Xiaoming
Liu, W.
Zhou, Guanglu
Wang, Y.
Lai, J.
A fast ℓ1-solver and its applications to robust face recognition
title A fast ℓ1-solver and its applications to robust face recognition
title_full A fast ℓ1-solver and its applications to robust face recognition
title_fullStr A fast ℓ1-solver and its applications to robust face recognition
title_full_unstemmed A fast ℓ1-solver and its applications to robust face recognition
title_short A fast ℓ1-solver and its applications to robust face recognition
title_sort fast ℓ1-solver and its applications to robust face recognition
url http://hdl.handle.net/20.500.11937/10229