Lightweight Image Super-Resolution with Adaptive Weighted Learning Network

TL;DR

This paper introduces AWSRN, a lightweight deep learning model for single-image super-resolution that balances high performance with low computational cost, suitable for real-world applications.

Contribution

The paper proposes a novel lightweight SR network with a local fusion block and adaptive multi-scale module, improving efficiency and performance over existing models.

Findings

Achieves superior super-resolution quality on standard datasets.

Maintains high performance with fewer parameters and lower computational cost.

Effective multi-scale feature utilization enhances reconstruction quality.

Abstract

Deep learning has been successfully applied to the single-image super-resolution (SISR) task with great performance in recent years. However, most convolutional neural network based SR models require heavy computation, which limit their real-world applications. In this work, a lightweight SR network, named Adaptive Weighted Super-Resolution Network (AWSRN), is proposed for SISR to address this issue. A novel local fusion block (LFB) is designed in AWSRN for efficient residual learning, which consists of stacked adaptive weighted residual units (AWRU) and a local residual fusion unit (LRFU). Moreover, an adaptive weighted multi-scale (AWMS) module is proposed to make full use of features in reconstruction layer. AWMS consists of several different scale convolutions, and the redundancy scale branch can be removed according to the contribution of adaptive weights in AWMS for lightweight network. The experimental results on the commonly used datasets show that the proposed lightweight AWSRN achieves superior performance on x2, x3, x4, and x8 scale factors to state-of-the-art methods with similar parameters and computational overhead. Code is avaliable at: https://github.com/ChaofWang/AWSRN