Abstract:In image denoising, deep convolutional neural networks (CNNs) can obtain favorable performance on removing spatially invariant noise. However, many of these networks cannot perform well on removing the real noise (i.e. spatially variant noise) generated during image acquisition or transmission, which severely sets back their application in practical image denoising tasks. Instead of continuously increasing the network depth, many researchers have revealed that expanding the width of networks can also be a useful way to improve model performance. It also has been verified that feature filtering can promote the learning ability of the models. Therefore, in this paper, we propose a novel Dual-branch Residual Attention Network (DRANet) for image denoising, which has both the merits of a wide model architecture and attention-guided feature learning. The proposed DRANet includes two different parallel branches, which can capture complementary features to enhance the learning ability of the model. We designed a new residual attention block (RAB) and a novel hybrid dilated residual attention block (HDRAB) for the upper and the lower branches, respectively. The RAB and HDRAB can capture rich local features through multiple skip connections between different convolutional layers, and the unimportant features are dropped by the residual attention modules. Meanwhile, the long skip connections in each branch, and the global feature fusion between the two parallel branches can capture the global features as well. Moreover, the proposed DRANet uses downsampling operations and dilated convolutions to increase the size of the receptive field, which can enable DRANet to capture more image context information. Extensive experiments demonstrate that compared with other state-of-the-art denoising methods, our DRANet can produce competitive denoising performance both on synthetic and real-world noise removal.

A Non-Local Enhanced Network for Image Restoration.

Residual Non-local Attention Networks for Image Restoration

Learning Deformable and Attentive Network for image restoration

Dilated Residual Encode-Decode Networks for Image Denoising

Residual Dense Network for Image Restoration

PNEN: Pyramid Non-Local Enhanced Networks

Local to non-local: Multi-scale progressive attention network for image restoration

LIGHTWEIGHT NON-LOCAL NETWORK FOR IMAGE SUPER-RESOLUTION

Image Restoration Based on End-to-End Unrolled Network

Image Denoising with Local Dense and Adaptive Global Residual Networks

LAE-Net: A Locally-Adaptive Emb E Dding Network for Low-Light Image Enhancement

Dual Residual Attention Network for Image Denoising

Combining Multi-Level Loss For Image Denoising

NHNet: A non-local hierarchical network for image denoising

Non-Local Auto-Encoder With Collaborative Stabilization for Image Restoration.

Enhanced CNN for Image Denoising

Hybrid Dilated Convolution with Attention Mechanisms for Image Denoising

Low-Illumination Image Enhancement Based on End-to-End Network Using Attention Module

DRCNet: Dynamic Image Restoration Contrastive Network.

NERNet: Noise Estimation and Removal Network for Image Denoising

Real Image Restoration via Structure-preserving Complementarity Attention