No abstract

No abstract

No abstract

No abstract

Learning based face hallucination methods have received much attention in recent years. As opposed to the existing methods, where the input image (patch) matrix is first converted into vectors before combination coefficients calculation, this paper proposes a novel matrix based regression model for directly combination coefficients calculation to preserve the structural information of the input matrix. For each lowresolution local patch matrix, its combination coefficients over the same position image patch matrices in training images can be computed. Then the corresponding high-resolution patch matrix can be obtained. Experiments conducted on the FERET face dataset indicate that our method could outperform other state-ofthe-art algorithms in terms of both vision and quantity.