Search

Abstract: This paper develops a novel framework that is capable of dealing with small sample size problem posed to subspace analysis methods for face representation and recognition. In the proposed framework, three aspects are presented. The first is the proposal of an iterative sampling technique. The second is adopting divide–conquer–merge strategy to incorporate the iterative sampling technique and subspace analysis method. The third is that the essence of 2D PCA is further explored. Experiments show that the proposed algorithm outperforms the traditional algorithms. [Copyright &y& Elsevier]

In this paper, linear local tangent space alignment (LLTSA), as a novel linear dimensionality reduction algorithm, is proposed. It uses the tangent space in the neighborhood of a data point to represent the local geometry, and then aligns those local tangent spaces in the low-dimensional space which is linearly mapped from the raw high-dimensional space. Since images of faces often belong to a manifold of intrinsically low dimension, we develop LLTSA algorithm for effective face manifold learning and recognition. Comprehensive comparisons and extensive experiments show that LLTSA achieves much higher recognition rates than a few competing methods. [Copyright &y& Elsevier]