LNMF-based Image Reconstruction for Face Recognition

Traditional LNMF-based face recognition methods obtain a universal subspace by using all trained images. The subspace mainly represents the commonness of human faces; there are a few sights of the individuality owned by a single person´s face. In this paper, we present a novel method named LNMF-based image reconstruction for face recognition. The fundamental difference between the traditional LNMF and ours is that we obtain the basis images by using each person´s pictures respectively, while the traditional way uses the whole training images of the database. After the step above, we obtained the features which will be employed to reconstruct the images by mapping the test images to the basis images. And then we adopt the minimum reconstruction error to finish the facial recognition. Finally, experiments based on the ORL database is illustrated the effectivity of this method.