Deeply-Recursive Convolutional Network for Image Super-Resolution

We propose an image super-resolution method (SR) using a deeply-recursive convolutional network (DRCN). Our network has a very deep recursive layer (up to 16 recursions). Increasing recursion depth can improve performance without introducing new parameters for additional convolutions. Albeit advantages, learning a DRCN is very hard with a standard gradient descent method due to exploding/ vanishing gradients. To ease the difficulty of training, we propose two extensions: recursive-supervision and skip-connection. Our method outperforms previous methods by a large margin.

1. Introduction

For image super-resolution (SR), receptive field of a convolutional network determines the amount of contextual information that can be exploited to infer missing high-frequency components. For example, if there exists a pattern with smoothed edges contained in a receptive field, it is plausible that the pattern is recognized and edges are appropriately sharpened. As SR is an ill-posed inverse problem, collecting and analyzing more neighbor pixels can possibly give more clues on what may be lost by downsampling.