Single Image Super-Resolution Method Using CNN-Based Lightweight Neural Networks

Seonjae Kim,Hunjoo Lee,Byung-Gyu Kim,Dongsan Jun,Eunjun Rhee

doi:10.3390/app11031092

Abstract

There are many studies that seek to enhance a low resolution image to a high resolution image in the area of super-resolution. As deep learning technologies have recently shown impressive results on the image interpolation and restoration field, recent studies are focusing on convolutional neural network (CNN)-based super-resolution schemes to surpass the conventional pixel-wise interpolation methods. In this paper, we propose two lightweight neural networks with a hybrid residual and dense connection structure to improve the super-resolution performance. In order to design the proposed networks, we extracted training images from the DIVerse 2K (DIV2K) image dataset and investigated the trade-off between the quality enhancement performance and network complexity under the proposed methods. The experimental results show that the proposed methods can significantly reduce both the inference speed and the memory required to store parameters and intermediate feature maps, while maintaining similar image quality compared to the previous methods.

Highlights

While the resolution of images has been rapidly increasing in recent years with the development of high-performance cameras, advanced image compression, and display panels, the demands to generate high resolution images from pre-existing low-resolution images are increasing for rendering on high resolution displays
We propose two SR-based lightweight neural networks (LNNs) with hybrid residual and dense networks, which are the “inter-layered SR-LNN” and “simplified SR-LNN”, respectively, which we denote in this paper as “SR-ILLNN” and “SR-SLNN”, respectively
The remainder of this paper is organized as follows: In Section 2, we review previous studies related to convolutional neural network (CNN)-based single image super-resolution (SISR) methods

Summary

Introduction

While the resolution of images has been rapidly increasing in recent years with the development of high-performance cameras, advanced image compression, and display panels, the demands to generate high resolution images from pre-existing low-resolution images are increasing for rendering on high resolution displays. Since the low-resolution images cannot represent the high-frequency information properly, most super-resolution (SR). Methods have focused on restoring high-frequency components. For this reason, SR methods are used to restore the high-frequency components from quantized images at the image and video post-processing stage [1,2,3]. SR methods are used to restore the high-frequency components from quantized images at the image and video post-processing stage [1,2,3] Deep learning schemes such as convolutional neural network (CNN) and multi-layer perceptron (MLP) are a branch of machine learning which aims to learn the correlations between input and output data. We propose two SR-based lightweight neural networks (LNNs) with hybrid residual and dense networks, which are the “inter-layered SR-LNN” and “simplified SR-LNN”, respectively, which we denote in this paper as “SR-ILLNN” and “SR-SLNN”, respectively

Methods

Results

Conclusion