Locality-Constrained Structral Orthogonal Procrustes Regression for Low-Resolution Face Recognition with Pose Variations

Abstract

Face images captured by surveillance cameras usually have low-resolution (LR) as well as uncontrolled poses and illumination conditions, which adversely affect the performance of face recognition algorithms. In this paper, we propose a locality-constrained structural orthogonal Procrustes regression (LCSOPR) approach to learn the pose-robust discriminative representations between LR and high-resolution (HR) images. The orthogonal Procrustes problem (OPP) seeks an optimal transformation between two images to correct the pose from one to the other. Additionally, our LCSOPR uses the nuclear norm constraint on the error term to keep image's structural information. Moreover, a locality constraint is also introduced to preserve the locality and the sparsity simultaneously. Finally, after getting the resolution-robust features, a simple yet powerful sparse representation based classifier engine is used to predict the face labels. The experimental results have shown that the proposed method can give better performance than many state-of-the-art LR face recognition approaches.