Efficient linear discriminant analysis with locality preserving for face recognition

Linear discriminant analysis (LDA) is one of the most popular techniques for extracting features in face recognition. LDA captures the global geometric structure. However, local geometric structure has recently been shown to be effective for face recognition. In this paper, we propose a novel featur...

Full description

Saved in:
Bibliographic Details
Published in:Pattern recognition Vol. 45; no. 5; pp. 1892 - 1898
Main Authors: Shu, Xin, Gao, Yao, Lu, Hongtao
Format: Journal Article
Language:English
Published: Kidlington Elsevier Ltd 01.05.2012
Elsevier
Subjects:
Algorithms
Applied sciences
Detection, estimation, filtering, equalization, prediction
Discriminant analysis
Exact sciences and technology
Face recognition
Image processing
Information, signal and communications theory
Least squares method
Linear discriminant analysis
Locality preserving projection
Mathematical models
Pattern recognition
Regression
Signal and communications theory
Signal processing
Signal, noise
Spectra
Spectral regression
Telecommunications and information theory
Tuning
Locality preserving projection
Linear discriminant analysis
Face recognition
Spectral regression
Biometrics
Singularity
Discriminant analysis
Image processing
Model selection
Least squares problem
Pattern recognition
Algorithm
Global local method
Regularization method
Algorithm performance
Signal processing
Feature extraction
Open market
Automatic recognition
ISSN:0031-3203, 1873-5142
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Linear discriminant analysis (LDA) is one of the most popular techniques for extracting features in face recognition. LDA captures the global geometric structure. However, local geometric structure has recently been shown to be effective for face recognition. In this paper, we propose a novel feature extraction algorithm which integrates both global and local geometric structures. We first cast LDA as a least square problem based on the spectral regression, then regularization technique is used to model the global and local geometric structures. Furthermore, we impose penalty on parameters to tackle the singularity problem and design an efficient model selection algorithm to choose the optimal tuning parameter which balances the tradeoff between the global and local structures. Experimental results on four well-known face data sets show that the proposed integration framework is competitive with traditional face recognition algorithms, which use either global or local structure only. ► We proposed a regularized least squares LDA which integrates both the global and local structures for face recognition. ► The formulation of regularized least squares LDA is based on spectral regression. ► The local structure is modeled via a regularization term defined by the graph Laplacian. ► We design an efficient algorithm for the estimation of the optimal tuning parameter.
Bibliography:ObjectType-Article-2
SourceType-Scholarly Journals-1
ObjectType-Feature-1
content type line 23
ISSN:0031-3203
1873-5142
DOI:10.1016/j.patcog.2011.11.012