Regularized locality preserving discriminant embedding for face recognition

Ying Han Pang, Beng Jin Teoh, Fazly Salleh Abas

Research output: Contribution to journalArticle

11 Citations (Scopus)

Abstract

For face recognition, graph embedding techniques attempt to produce a high data locality projection for better recognition performance. However, estimation of population data locality could be severely biased due to small number of training samples. The biased estimation triggers overfitting problem and hence poor generalization. In this paper, we propose a new linear graph embedding technique based upon an adaptive locality preserving regulation model (ALPRM), known as Regularized Locality Preserving Discriminant Embedding (RLPDE). In RLPDE, the projection features are regulated based on ALPRM to approach population data locality, which can directly enhance the locality preserving capability of the projection features. This paper also presents the relation between locality preserving capability and class discrimination. Specifically, we show that the optimization of the locality preserving function minimizes the within-class variability. Experiments on three face datasets such as PIE, FRGC and FERET show the promising performance of the proposed technique.

Original languageEnglish
Pages (from-to)156-166
Number of pages11
JournalNeurocomputing
Volume77
Issue number1
DOIs
Publication statusPublished - 2012 Feb 1

Fingerprint

Face recognition
Population
Experiments
Facial Recognition
Datasets

All Science Journal Classification (ASJC) codes

  • Computer Science Applications
  • Cognitive Neuroscience
  • Artificial Intelligence

Cite this

Pang, Ying Han ; Teoh, Beng Jin ; Abas, Fazly Salleh. / Regularized locality preserving discriminant embedding for face recognition. In: Neurocomputing. 2012 ; Vol. 77, No. 1. pp. 156-166.
@article{3b4e80e7324b45d8a54afcc21633ea3c,
title = "Regularized locality preserving discriminant embedding for face recognition",
abstract = "For face recognition, graph embedding techniques attempt to produce a high data locality projection for better recognition performance. However, estimation of population data locality could be severely biased due to small number of training samples. The biased estimation triggers overfitting problem and hence poor generalization. In this paper, we propose a new linear graph embedding technique based upon an adaptive locality preserving regulation model (ALPRM), known as Regularized Locality Preserving Discriminant Embedding (RLPDE). In RLPDE, the projection features are regulated based on ALPRM to approach population data locality, which can directly enhance the locality preserving capability of the projection features. This paper also presents the relation between locality preserving capability and class discrimination. Specifically, we show that the optimization of the locality preserving function minimizes the within-class variability. Experiments on three face datasets such as PIE, FRGC and FERET show the promising performance of the proposed technique.",
author = "Pang, {Ying Han} and Teoh, {Beng Jin} and Abas, {Fazly Salleh}",
year = "2012",
month = "2",
day = "1",
doi = "10.1016/j.neucom.2011.09.007",
language = "English",
volume = "77",
pages = "156--166",
journal = "Neurocomputing",
issn = "0925-2312",
publisher = "Elsevier",
number = "1",

}

Regularized locality preserving discriminant embedding for face recognition. / Pang, Ying Han; Teoh, Beng Jin; Abas, Fazly Salleh.

In: Neurocomputing, Vol. 77, No. 1, 01.02.2012, p. 156-166.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Regularized locality preserving discriminant embedding for face recognition

AU - Pang, Ying Han

AU - Teoh, Beng Jin

AU - Abas, Fazly Salleh

PY - 2012/2/1

Y1 - 2012/2/1

N2 - For face recognition, graph embedding techniques attempt to produce a high data locality projection for better recognition performance. However, estimation of population data locality could be severely biased due to small number of training samples. The biased estimation triggers overfitting problem and hence poor generalization. In this paper, we propose a new linear graph embedding technique based upon an adaptive locality preserving regulation model (ALPRM), known as Regularized Locality Preserving Discriminant Embedding (RLPDE). In RLPDE, the projection features are regulated based on ALPRM to approach population data locality, which can directly enhance the locality preserving capability of the projection features. This paper also presents the relation between locality preserving capability and class discrimination. Specifically, we show that the optimization of the locality preserving function minimizes the within-class variability. Experiments on three face datasets such as PIE, FRGC and FERET show the promising performance of the proposed technique.

AB - For face recognition, graph embedding techniques attempt to produce a high data locality projection for better recognition performance. However, estimation of population data locality could be severely biased due to small number of training samples. The biased estimation triggers overfitting problem and hence poor generalization. In this paper, we propose a new linear graph embedding technique based upon an adaptive locality preserving regulation model (ALPRM), known as Regularized Locality Preserving Discriminant Embedding (RLPDE). In RLPDE, the projection features are regulated based on ALPRM to approach population data locality, which can directly enhance the locality preserving capability of the projection features. This paper also presents the relation between locality preserving capability and class discrimination. Specifically, we show that the optimization of the locality preserving function minimizes the within-class variability. Experiments on three face datasets such as PIE, FRGC and FERET show the promising performance of the proposed technique.

UR - http://www.scopus.com/inward/record.url?scp=80955134088&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=80955134088&partnerID=8YFLogxK

U2 - 10.1016/j.neucom.2011.09.007

DO - 10.1016/j.neucom.2011.09.007

M3 - Article

VL - 77

SP - 156

EP - 166

JO - Neurocomputing

JF - Neurocomputing

SN - 0925-2312

IS - 1

ER -