Factored principal components analysis, with applications to face recognition

Ian L. Dryden; Li Bai; Christopher J. Brignell; Linlin Shen

doi:10.1007/s11222-008-9087-6

Factored principal components analysis, with applications to face recognition

Ian L. Dryden, Li Bai, Christopher J. Brignell, Linlin Shen

Research output: Journal Publication › Article › peer-review

13 Citations (Scopus)

Abstract

A dimension reduction technique is proposed for matrix data, with applications to face recognition from images. In particular, we propose a factored covariance model for the data under study, estimate the parameters using maximum likelihood, and then carry out eigendecompositions of the estimated covariance matrix. We call the resulting method factored principal components analysis. We also develop a method for classification using a likelihood ratio criterion, which has previously been used for evaluating the strength of forensic evidence. The methodology is illustrated with applications in face recognition.

Original language	English
Pages (from-to)	229-238
Number of pages	10
Journal	Statistics and Computing
Volume	19
Issue number	3
DOIs	https://doi.org/10.1007/s11222-008-9087-6
Publication status	Published - Sept 2009
Externally published	Yes

Keywords

Face recognition
Forensic identification
Gabor wavelets
Kernel density estimator
Likelihood ratio
Multivariate normal
Principal components analysis

ASJC Scopus subject areas

Theoretical Computer Science
Statistics and Probability
Statistics, Probability and Uncertainty
Computational Theory and Mathematics

Access to Document

10.1007/s11222-008-9087-6

Cite this

@article{c3ea282c6b2c4830ace48a8e0fe07367,

title = "Factored principal components analysis, with applications to face recognition",

abstract = "A dimension reduction technique is proposed for matrix data, with applications to face recognition from images. In particular, we propose a factored covariance model for the data under study, estimate the parameters using maximum likelihood, and then carry out eigendecompositions of the estimated covariance matrix. We call the resulting method factored principal components analysis. We also develop a method for classification using a likelihood ratio criterion, which has previously been used for evaluating the strength of forensic evidence. The methodology is illustrated with applications in face recognition.",

keywords = "Face recognition, Forensic identification, Gabor wavelets, Kernel density estimator, Likelihood ratio, Multivariate normal, Principal components analysis",

author = "Dryden, {Ian L.} and Li Bai and Brignell, {Christopher J.} and Linlin Shen",

note = "Funding Information: Acknowledgements The research was supported by an EPSRC Discipline Hopping Grant EP/C549066/1. We are grateful to Angel Serrano for discussions on B2DPCA, and two anonymous referees for their helpful comments.",

year = "2009",

month = sep,

doi = "10.1007/s11222-008-9087-6",

language = "English",

volume = "19",

pages = "229--238",

journal = "Statistics and Computing",

issn = "0960-3174",

publisher = "Springer Netherlands",

number = "3",

}

TY - JOUR

T1 - Factored principal components analysis, with applications to face recognition

AU - Dryden, Ian L.

AU - Bai, Li

AU - Brignell, Christopher J.

AU - Shen, Linlin

N1 - Funding Information: Acknowledgements The research was supported by an EPSRC Discipline Hopping Grant EP/C549066/1. We are grateful to Angel Serrano for discussions on B2DPCA, and two anonymous referees for their helpful comments.

PY - 2009/9

Y1 - 2009/9

N2 - A dimension reduction technique is proposed for matrix data, with applications to face recognition from images. In particular, we propose a factored covariance model for the data under study, estimate the parameters using maximum likelihood, and then carry out eigendecompositions of the estimated covariance matrix. We call the resulting method factored principal components analysis. We also develop a method for classification using a likelihood ratio criterion, which has previously been used for evaluating the strength of forensic evidence. The methodology is illustrated with applications in face recognition.

AB - A dimension reduction technique is proposed for matrix data, with applications to face recognition from images. In particular, we propose a factored covariance model for the data under study, estimate the parameters using maximum likelihood, and then carry out eigendecompositions of the estimated covariance matrix. We call the resulting method factored principal components analysis. We also develop a method for classification using a likelihood ratio criterion, which has previously been used for evaluating the strength of forensic evidence. The methodology is illustrated with applications in face recognition.

KW - Face recognition

KW - Forensic identification

KW - Gabor wavelets

KW - Kernel density estimator

KW - Likelihood ratio

KW - Multivariate normal

KW - Principal components analysis

UR - http://www.scopus.com/inward/record.url?scp=67650680897&partnerID=8YFLogxK

U2 - 10.1007/s11222-008-9087-6

DO - 10.1007/s11222-008-9087-6

M3 - Article

AN - SCOPUS:67650680897

SN - 0960-3174

VL - 19

SP - 229

EP - 238

JO - Statistics and Computing

JF - Statistics and Computing

IS - 3

ER -

Factored principal components analysis, with applications to face recognition

Abstract

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this