An improvement in feature feedback using R-LDA with application to Yale database

Authors:
Lang Bach Truong;Sang-Il Choi;Gu-Min Jeong;Jeong-Min Seo
Affiliations:
School of Electrical Engineering, Kookmin University, Seoul, Korea;Dept. of Computer Science, University of Southern California;School of Electrical Engineering, Kookmin University, Seoul, Korea;Dept. of Computer Engineering, Konkuk University, Seoul, Korea
Venue:
ICHIT'11 Proceedings of the 5th international conference on Convergence and hybrid information technology
Year:
2011

Citing 7
Cited 0

Introduction to statistical pattern recognition (2nd ed.)

Introduction to statistical pattern recognition (2nd ed.)
Eigenfaces vs. Fisherfaces: Recognition Using Class Specific Linear Projection

IEEE Transactions on Pattern Analysis and Machine Intelligence
Statistical Pattern Recognition: A Review

IEEE Transactions on Pattern Analysis and Machine Intelligence
Regularization studies of linear discriminant analysis in small sample size scenarios with application to face recognition

Pattern Recognition Letters
Eigenfaces for recognition

Journal of Cognitive Neuroscience
A novel Gabor-LDA based face recognition method

PCM'04 Proceedings of the 5th Pacific Rim conference on Advances in Multimedia Information Processing - Volume Part I
Face recognition using LDA-based algorithms

IEEE Transactions on Neural Networks

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper improves the performance of Feature Feedback and presents its application to face recognition. Feature Feedback has been introduced as a method which focuses on preprocessing the input data before classification. After extracting the features from original, Feature Feedback identifies the important part of the original data through the reverse mapping from the extracted features to the original space. In the feature extraction step, original feature feedback used PCA before LDA to avoid the small sample size problem but it has been shown that this may cause loss of significant discriminatory information. To overcome that problem, in the proposed method, we introduce feature feedback using regularized Fisher's separability criterion to extract the features and apply it to face recognition using the Yale data. The experimental results show that the proposed method works well.