Paper Info
Title
Optimizing principal component analysis performance for face recognition using genetic algorithm
Abstract
Principal Component Analysis (PCA) turns out to be one of the most successful techniques in face recognition systems as a statistical method for dimensionality reduction. Even so, it is yet not optimal from the perspective of classification because the underlying distribution among different face classes in the image space is unpredicted and not known in advance. Besides, in practical applications, a question always raised on how much data should be included in the training. In this paper, a technique that associates genetic algorithm (GA) to PCA is proposed to maintain the property of PCA while enhancing the classification performance. It reconsiders the available training data and tries to find the best underlying distribution for classification. ORL, and Yale A databases have been used in the experiments to analyze and evaluate the performance of the proposed method compared to original PCA. The experiment results reveal that the proposed method outperforms PCA in terms of accuracy and classification time.
Year
DOI
Venue
2014
10.1016/j.neucom.2013.08.022
Neurocomputing
Keywords
Field
DocType
different face class,classification time,underlying distribution,genetic algorithm,face recognition system,principal component analysis,original pca,available training data,principal component analysis performance,statistical method,classification performance,pca,face recognition
Training set,Facial recognition system,Dimensionality reduction,Pattern recognition,Computer science,Artificial intelligence,Machine learning,Genetic algorithm,Principal component analysis
Journal
Volume
ISSN
Citations 
128,
0925-2312
7
PageRank 
References 
Authors
0.45
9
3
Name
Order
Citations
PageRank
Waled Hussein Al-Arashi1102.21
Haidi Ibrahim21158.83
Shahrel Azmin Suandi310512.72