Two-dimensional Linear discriminant analysis for low-resolution face recognition
Di Zhao, Zhenxue Chen, Chengyun Liu, Yanan Peng · 2017
Low-resolution (LR) is a challenging problem in the real world. In order to obtain better performance for low-resolution face recognition (LRFR), this paper employs a novel approach for matching low-resolution images with high resolution (HR) images based on two-dimensional linear discriminant analysis (2D-LDA) and metric learning method. The LR and HR images are transformed into a common space via the 2D-LDA method in which the most discriminative information between them is preserved. Also, it overcomes the singularity and loss of the spatial information problem because of its matrix representation. To further improve the recognition performance, metric learning method is used based on neighborhood component analysis (NCA) which aims to maximize the leave-one-out (LOO) classification accuracy. Experiments on the ORL database with a wide range of resolutions illustrate the usefulness of the proposed method.