L1-norm based linear discriminant analysis: An application to face recognition

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

Linear Discriminant Analysis (LDA) is a well-known feature extraction method for supervised subspace learning in statistical pattern recognition. In this paper, a novel method of LDA based on a new L1-norm optimization technique and its variances are proposed. The conventional LDA, which is based on L2-norm, is sensitivity to the presence of outliers, since it used the L2-norm to measure the between-class and within-class distances. In addition, the conventional LDA often suffers from the so-called small sample size (3S) problem since the number of samples is always smaller than the dimension of the feature space in many applications, such as face recognition. Based on L1-norm, the proposed methods have several advantages, first they are robust to outliers because they utilize the L1-norm, which is less sensitive to outliers. Second, they have no 3S problem. Third, they are invariant to rotations as well. The proposed methods are capable of reducing the influence of outliers substantially, resulting in a robust classification. Performance assessment in face application shows that the proposed approaches are more effectiveness to address outliers issue than traditional ones.

Original languageEnglish
Pages (from-to)550-558
Number of pages9
JournalIEICE Transactions on Information and Systems
VolumeE96-D
Issue number3
DOIs
Publication statusPublished - 2013 Mar

Fingerprint

Discriminant analysis
Face recognition
Supervised learning
Pattern recognition
Feature extraction

Keywords

  • 2DLDA-L1
  • BLDA-L1
  • Face recognition
  • L1-norm
  • LDA-L1
  • Linear discriminant analysis
  • Linear programming

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Software
  • Artificial Intelligence
  • Hardware and Architecture
  • Computer Vision and Pattern Recognition

Cite this

L1-norm based linear discriminant analysis : An application to face recognition. / Zhou, Wei; Kamata, Seiichiro.

In: IEICE Transactions on Information and Systems, Vol. E96-D, No. 3, 03.2013, p. 550-558.

Research output: Contribution to journalArticle

@article{6d0342a5aea64c7da01fd150cded6f6b,
title = "L1-norm based linear discriminant analysis: An application to face recognition",
abstract = "Linear Discriminant Analysis (LDA) is a well-known feature extraction method for supervised subspace learning in statistical pattern recognition. In this paper, a novel method of LDA based on a new L1-norm optimization technique and its variances are proposed. The conventional LDA, which is based on L2-norm, is sensitivity to the presence of outliers, since it used the L2-norm to measure the between-class and within-class distances. In addition, the conventional LDA often suffers from the so-called small sample size (3S) problem since the number of samples is always smaller than the dimension of the feature space in many applications, such as face recognition. Based on L1-norm, the proposed methods have several advantages, first they are robust to outliers because they utilize the L1-norm, which is less sensitive to outliers. Second, they have no 3S problem. Third, they are invariant to rotations as well. The proposed methods are capable of reducing the influence of outliers substantially, resulting in a robust classification. Performance assessment in face application shows that the proposed approaches are more effectiveness to address outliers issue than traditional ones.",
keywords = "2DLDA-L1, BLDA-L1, Face recognition, L1-norm, LDA-L1, Linear discriminant analysis, Linear programming",
author = "Wei Zhou and Seiichiro Kamata",
year = "2013",
month = "3",
doi = "10.1587/transinf.E96.D.550",
language = "English",
volume = "E96-D",
pages = "550--558",
journal = "IEICE Transactions on Information and Systems",
issn = "0916-8532",
publisher = "Maruzen Co., Ltd/Maruzen Kabushikikaisha",
number = "3",

}

TY - JOUR

T1 - L1-norm based linear discriminant analysis

T2 - An application to face recognition

AU - Zhou, Wei

AU - Kamata, Seiichiro

PY - 2013/3

Y1 - 2013/3

N2 - Linear Discriminant Analysis (LDA) is a well-known feature extraction method for supervised subspace learning in statistical pattern recognition. In this paper, a novel method of LDA based on a new L1-norm optimization technique and its variances are proposed. The conventional LDA, which is based on L2-norm, is sensitivity to the presence of outliers, since it used the L2-norm to measure the between-class and within-class distances. In addition, the conventional LDA often suffers from the so-called small sample size (3S) problem since the number of samples is always smaller than the dimension of the feature space in many applications, such as face recognition. Based on L1-norm, the proposed methods have several advantages, first they are robust to outliers because they utilize the L1-norm, which is less sensitive to outliers. Second, they have no 3S problem. Third, they are invariant to rotations as well. The proposed methods are capable of reducing the influence of outliers substantially, resulting in a robust classification. Performance assessment in face application shows that the proposed approaches are more effectiveness to address outliers issue than traditional ones.

AB - Linear Discriminant Analysis (LDA) is a well-known feature extraction method for supervised subspace learning in statistical pattern recognition. In this paper, a novel method of LDA based on a new L1-norm optimization technique and its variances are proposed. The conventional LDA, which is based on L2-norm, is sensitivity to the presence of outliers, since it used the L2-norm to measure the between-class and within-class distances. In addition, the conventional LDA often suffers from the so-called small sample size (3S) problem since the number of samples is always smaller than the dimension of the feature space in many applications, such as face recognition. Based on L1-norm, the proposed methods have several advantages, first they are robust to outliers because they utilize the L1-norm, which is less sensitive to outliers. Second, they have no 3S problem. Third, they are invariant to rotations as well. The proposed methods are capable of reducing the influence of outliers substantially, resulting in a robust classification. Performance assessment in face application shows that the proposed approaches are more effectiveness to address outliers issue than traditional ones.

KW - 2DLDA-L1

KW - BLDA-L1

KW - Face recognition

KW - L1-norm

KW - LDA-L1

KW - Linear discriminant analysis

KW - Linear programming

UR - http://www.scopus.com/inward/record.url?scp=84878234653&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84878234653&partnerID=8YFLogxK

U2 - 10.1587/transinf.E96.D.550

DO - 10.1587/transinf.E96.D.550

M3 - Article

AN - SCOPUS:84878234653

VL - E96-D

SP - 550

EP - 558

JO - IEICE Transactions on Information and Systems

JF - IEICE Transactions on Information and Systems

SN - 0916-8532

IS - 3

ER -