Face recognition based on two-phase 2DLEM

Some research efforts have shown that face images possibly reside on a nonlinear sub-manifold. This paper proposes a new feature extraction method, Two Dimensional Laplacian Eigen-Map, (2DLEM), which especially considers the manifold structures of the face images, and extracts the proper features from image matrix directly by using a linear transformation. 2DLEM has a disadvantage that it requires more coefficients for image representation as 2D feature extraction algorithms have. In this paper, we propose a two-phase framework to overcome the disadvantage. The experiments on the ORL database show that 2DLEM achieves the best performance, and the two-phase framework can reduce the dimension effectively.