Deep Networks for Image Super-Resolution with Sparse Prior

TL;DR

This paper combines domain expertise from sparse coding with deep neural networks to improve image super-resolution, achieving better accuracy and subjective quality than existing methods.

Contribution

It introduces a neural network model inspired by sparse coding specifically designed for super-resolution, trained end-to-end with improved efficiency and reduced size.

Findings

Outperforms state-of-the-art super-resolution methods in accuracy

Achieves higher subjective quality in image restoration

Demonstrates the effectiveness of integrating sparse coding with deep learning

Abstract

Deep learning techniques have been successfully applied in many areas of computer vision, including low-level image restoration problems. For image super-resolution, several models based on deep neural networks have been recently proposed and attained superior performance that overshadows all previous handcrafted models. The question then arises whether large-capacity and data-driven models have become the dominant solution to the ill-posed super-resolution problem. In this paper, we argue that domain expertise represented by the conventional sparse coding model is still valuable, and it can be combined with the key ingredients of deep learning to achieve further improved results. We show that a sparse coding model particularly designed for super-resolution can be incarnated as a neural network, and trained in a cascaded structure from end to end. The interpretation of the network based on sparse coding leads to much more efficient and effective training, as well as a reduced model size. Our model is evaluated on a wide range of images, and shows clear advantage over existing state-of-the-art methods in terms of both restoration accuracy and human subjective quality.