Enhanced Deep Residual Networks for Single Image Super-Resolution

TL;DR

This paper introduces an enhanced deep residual network (EDSR) and a multi-scale super-resolution system (MDSR) that outperform existing methods in single image super-resolution, winning the NTIRE2017 challenge.

Contribution

The paper presents a novel EDSR architecture by removing unnecessary modules and expanding model size, along with a multi-scale system for multiple upscaling factors, advancing super-resolution techniques.

Findings

EDSR surpasses state-of-the-art super-resolution methods on benchmarks.

The multi-scale MDSR effectively reconstructs images at different upscaling factors.

The proposed methods won the NTIRE2017 Super-Resolution Challenge.

Abstract

Recent research on super-resolution has progressed with the development of deep convolutional neural networks (DCNN). In particular, residual learning techniques exhibit improved performance. In this paper, we develop an enhanced deep super-resolution network (EDSR) with performance exceeding those of current state-of-the-art SR methods. The significant performance improvement of our model is due to optimization by removing unnecessary modules in conventional residual networks. The performance is further improved by expanding the model size while we stabilize the training procedure. We also propose a new multi-scale deep super-resolution system (MDSR) and training method, which can reconstruct high-resolution images of different upscaling factors in a single model. The proposed methods show superior performance over the state-of-the-art methods on benchmark datasets and prove its excellence by winning the NTIRE2017 Super-Resolution Challenge.