Optimized projection for sparse representation based classification

Abstract

Dimensionality reduction(DR) methods have commonly been used as a principled way to understand the high-dimensional data such as face images. In this paper, we propose a new supervised DR method called Optimized Projection for Sparse Representation based Classification(OP-SRC). SRC assumes that any new sample will approximately lie in the linear span of the training samples sharing the same class label. The decision of SRC is based on the reconstruction residual. OP-SRC aims to reduce the within-class reconstruction residual and simultaneously increases the between-class reconstruction residual. Therefore, SRC performs well in the OP-SRC transformed space. The feasibility and effectiveness of the proposed method is verified on Yale and ORL with promising results. © 2011 Springer-Verlag.