Title :
Sparsity Preserving-Based Local Fisher Discriminant Analysis with Applications in Face Recognition
Author_Institution :
Coll. of Math. & Comput. Sci., Qinzhou Coll., Qinzhou, China
Abstract :
A kind of algorithm called sparsity preserving-based local fisher discriminant analysis (SPLFDA) is proposed, which insulates sparsity preserving projections and local fisher discriminant analysis in the process of dimensionality reduction. It inherits the special character of geometrical structure preserving and neighborhood preserving. Experiments operated on UMIST, Yale and YaleB face dataset show that the algorithm is more effective.
Keywords :
face recognition; statistical analysis; visual databases; SPLFDA; UMIST dataset; Yale dataset; YaleB face dataset; dimensionality reduction; face recognition; geometrical structure preservation; neighborhood preservation; sparsity preserving projections; sparsity preserving-based local Fisher discriminant analysis; Accuracy; Algorithm design and analysis; Classification algorithms; Databases; Face; Face recognition; Training; face recognition; local fisher discriminant analysis; semi-supervised dimensionality reduction; sparsity preserving;
Conference_Titel :
Computational and Information Sciences (ICCIS), 2012 Fourth International Conference on
Conference_Location :
Chongqing
Print_ISBN :
978-1-4673-2406-9
DOI :
10.1109/ICCIS.2012.289
