GUN: Gradual Upsampling Network for Single Image Super-Resolution

TL;DR

GUN introduces a gradual upsampling approach for single image super-resolution using deep CNNs, simplifying the learning process and improving the quality of reconstructed high-resolution images.

Contribution

The paper proposes a novel gradual upsampling network with a training strategy that enhances super-resolution performance and training efficiency.

Findings

Effective super-resolution results on multiple datasets.

Simplified training process with gradual learning.

Improved image detail and vividness in reconstructions.

Abstract

In this paper, an efficient super-resolution (SR) method based on deep convolutional neural network (CNN) is proposed, namely Gradual Upsampling Network (GUN). Recent CNN based SR methods often preliminarily magnify the low resolution (LR) input to high resolution (HR) and then reconstruct the HR input, or directly reconstruct the LR input and then recover the HR result at the last layer. The proposed GUN utilizes a gradual process instead of these two commonly used frameworks. The GUN consists of an input layer, multiple upsampling and convolutional layers, and an output layer. By means of the gradual process, the proposed network can simplify the direct SR problem to multistep easier upsampling tasks with very small magnification factor in each step. Furthermore, a gradual training strategy is presented for the GUN. In the proposed training process, an initial network can be easily trained with edge-like samples, and then the weights are gradually tuned with more complex samples. The GUN can recover fine and vivid results, and is easy to be trained. The experimental results on several image sets demonstrate the effectiveness of the proposed network.