Abstract:For the deep learning based super-resolution (SR) reconstruction method, researchers try to expand the receptive field to improve the reconstruction quality. With the increase of network depth, the consumption of computing resources is impressive. Therefore, SR technology is challenging to apply to small mobile clients. To deal with this defect, we propose a lightweight multi-stage residual distillation network (MRDN) for the SR task in this paper. The model has made the following two main improvements: First, we design a multi-stage residual distillation block (MRDB). It combines the channel separation and the skip connection to reduce the parameter number and guarantee the network’s performance by residual learning. Secondly, we propose efficient pixel attention (EPA) module, which weights different channels according to their importance so that the network can learn more details of pictures. Experiments show that the proposed MRDN is superior to the state-of-the-art models in subjective visual effects and objective evaluation criteria such as PSNR, SSIM, and computational complexity. Taking the basic data set manga109 as an example, the PSNR of this algorithm on scale ×4 reaches 30.66, and the average inference time is 0.021 s. The performance surpasses RFDN, the champion of AIM 2020 Challenge on Efficient Super-Resolution.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is: **How to design a lightweight multi - stage residual distillation network (MRDN) to significantly reduce the consumption of computational resources while ensuring the quality of image super - resolution reconstruction, so that it can be applied to devices with limited computing power such as mobile clients?** Specifically, the paper addresses the following problems in the image super - resolution (SR) task in deep learning: 1. **High consumption of computational resources**: As the network depth increases, traditional SR models require a large amount of computational resources, which limits their application on small clients such as mobile devices. 2. **High model complexity**: Many high - performance SR models (such as RCAN, EDSR, etc.) have a large number of parameters and complex structures, resulting in long training and inference times. 3. **Balancing reconstruction quality and model complexity**: How to reduce the complexity and computational cost of the model while maintaining high - quality image reconstruction effects. To solve these problems, the paper proposes a lightweight multi - stage residual distillation network (MRDN), which achieves the goal mainly through the following innovations: - Design a multi - stage residual distillation block (MRDB), combine channel separation and skip connections to reduce the number of parameters and ensure network performance. - Propose an efficient pixel - attention module (EPA), weight pixels according to their importance, highlight important features, and maintain a low computational overhead at the same time. The experimental results show that the proposed MRDN model is superior to the existing lightweight SR models in terms of subjective visual effects and objective evaluation metrics (such as PSNR, SSIM, and computational complexity), and achieves a PSNR value of 30.66 at a ×4 scale on the manga109 dataset, with an average inference time of 0.021 seconds, outperforming the champion model RFDN of the AIM 2020 Efficient Super - Resolution Challenge.

MRDN: A lightweight Multi-stage residual distillation network for image Super-Resolution

Residual multi-branch distillation network for efficient image super-resolution

Multiscale Residual Dense Network for the Super-Resolution of Remote Sensing Images

Multi-scale dense spatially-adaptive residual distillation network for lightweight underwater image super-resolution

Lightweight multi-scale distillation attention network for image super-resolution

A Multi-Attention Feature Distillation Neural Network for Lightweight Single Image Super-Resolution

RIRGAN: an End-to-end Lightweight Multi-Task Learning Method for Brain MRI Super-Resolution and Denoising

MRA-IDN: A Lightweight Super-Resolution Framework of Remote Sensing Images based on Multi-Scale Residual Attention Fusion Mechanism

Multi-Source Deep Residual Fusion Network for Depth Image Super-resolution

Multi-Scale Deep Neural Networks for Real Image Super-Resolution

CFDN: cross-scale feature distillation network for lightweight single image super-resolution

MRFN: Multi-Receptive-Field Network for Fast and Accurate Single Image Super-Resolution

A Lightweight Image Super-Resolution Reconstruction Algorithm Based on the Residual Feature Distillation Mechanism

Residual Feature Distillation Network for Lightweight Image Super-Resolution

DDistill-SR: Reparameterized Dynamic Distillation Network for Lightweight Image Super-Resolution

Cascaded Deep Networks with Multiple Receptive Fields for Infrared Image Super-Resolution

An Efficient Feature Reuse Distillation Network for Lightweight Image Super-Resolution

Single-image Super-Resolution with Multilevel Residual Attention Network

Distilling with Residual Network for Single Image Super Resolution

DIPNet: Efficiency Distillation and Iterative Pruning for Image Super-Resolution

End-to-End Super-Resolution for Remote-Sensing Images Using an Improved Multi-Scale Residual Network