Face Recognition Using L-Fisherfaces
Cheng-Yuan Zhang,Qiu-Qi Ruan
DOI: https://doi.org/10.6688/jise.2010.26.4.23
2010-01-01
Journal of information science and engineering
Abstract:An appearance-based face recognition approach called the L-Fisherfaces is proposed in this paper, By using Local Fisher Discriminant Embedding (LFDE), the face images are mapped into a face subspace for analysis. Different from Linear Discriminant Analysis (LDA), which effectively sees only the Euclidean structure of face space, LFDE finds an em bedding that preserves local information, and obtains a face subspace that best detects the essential face manifold structure. Different from Locality Preserving Projections (LPP) and Unsupervised Discriminant projections (UDP), which ignore the class label information, LFDE searches for the project axes on which the data points of different classes are far from each other while requiring data points of the same class to be close to each other. W e compare the proposed L-Fisherfaces approach with PCA, LDA, LPP, and UDP on three different face databases. Experimental results suggest that the proposed L-Fisherfaces provides a better representation and achieves higher accuracy in face recognition.