Abstract:
The problem of determining the optimal set of discriminant vectors for feature extraction in pattern recognition is investigated. We propose a new direct LDA (D-LDA) meth...Show MoreMetadata
Abstract:
The problem of determining the optimal set of discriminant vectors for feature extraction in pattern recognition is investigated. We propose a new direct LDA (D-LDA) method that is applicable to small sample size (SSS) problems often arising in face recognition. The experimental results on two popular databases show the effectiveness of the proposed method.
Published in: Proceedings of the 17th International Conference on Pattern Recognition, 2004. ICPR 2004.
Date of Conference: 26-26 August 2004
Date Added to IEEE Xplore: 20 September 2004
Print ISBN:0-7695-2128-2
Print ISSN: 1051-4651
Contents 