DRSFANet: Dual-Path CNN with Residual and Frequency Attention for Image Denoising

Image noise, often resulting from disturbances during image acquisition or imperfections in the imaging device, notably degrades the quality of digital images. The challenge of removing this noise has been addressed through various techniques, from traditional filtering and prior-based methods to more recent deep learning approaches. In this paper, we introduce DRSFANet, an advanced dual-path convolutional neural network (CNN) specifically designed to tackle both synthetic Additive White Gaussian Noise (AWGN) and real-world noise. DRSFANet incorporates several innovative components: a residual feature extraction module (FEB) equipped with dilated convolutional layers to enhance the receptive field and mitigate gradient vanishing issues, and novel attention modules—Frequency-Plane Attention Block (FPAB) and Residual Attention Block (RAB)—which improve feature extraction in both frequency and spatial domains. Furthermore, the model features a downsampling (DS) block that effectively consolidates essential features prior to their integration into subsequent network stages. Comprehensive experimental evaluations reveal that DRSFANet outperforms several state-of-the-art denoising methods, demonstrating superior performance in both synthetic and real datasets through rigorous quantitative and qualitative analysis.

I. Introduction

Noise is a common phenomenon in all acquired images and is primarily introduced due to disturbances in the medium as well as due to measurement imperfections of the acquisition device. The most commonly occurring noise in digital images is the additive white Gaussian noise (AWGN), the presence of which degrades the quality of the acquired image as well as poses a limitation on any further processing of those images. The purpose of image denoising is to recover a noise-free clean image from the acquired degraded images. Image denoising by nature is an ill-posed problem and requires very strong prior knowledge for image restoration. Various strategies have been employed in the literature that ranges from filtering-based methods [1] to image prior-based methods (like Nonself similarity models [2], gradient models [3]), sparsity-based models [4], Markov random field models [5] to the more recent deep neural networks [6] based models.