Learning Multi-scale Spatial-frequency Features for Image Denoising

TL;DR

This paper introduces MADNet, a multi-scale adaptive dual-domain network that leverages image pyramids and a novel spatial-frequency learning unit to improve image denoising by effectively handling different frequency noise components.

Contribution

The paper proposes a novel MADNet architecture with an adaptive spatial-frequency learning unit and global feature fusion, addressing limitations of fixed architectures and uniform frequency treatment.

Findings

Outperforms current state-of-the-art denoising methods on synthetic datasets.

Effective separation of high-frequency and low-frequency noise components.

Demonstrates robustness on real noisy image datasets.

Abstract

Recent advancements in multi-scale architectures have demonstrated exceptional performance in image denoising tasks. However, existing architectures mainly depends on a fixed single-input single-output Unet architecture, ignoring the multi-scale representations of pixel level. In addition, previous methods treat the frequency domain uniformly, ignoring the different characteristics of high-frequency and low-frequency noise. In this paper, we propose a novel multi-scale adaptive dual-domain network (MADNet) for image denoising. We use image pyramid inputs to restore noise-free results from low-resolution images. In order to realize the interaction of high-frequency and low-frequency information, we design an adaptive spatial-frequency learning unit (ASFU), where a learnable mask is used to separate the information into high-frequency and low-frequency components. In the skip connections, we design a global feature fusion block to enhance the features at different scales. Extensive experiments on both synthetic and real noisy image datasets verify the effectiveness of MADNet compared with current state-of-the-art denoising approaches.