Abstract:Image denoising remains a classical yet still challenging problem, because the noise can destroy details and cause severe information loss. In recent years, various well-designed CNN-based methods have been extensively applied in image denoising because of the strong learning ability. However, most of them share an unidirectional procedure, which directly learns a mapping from noisy input to a clean image without focusing on the over-smoothed state of the denoising process, limiting the richness of extracted features. Different from previous works, we propose a blurred image feature guided CNN (BFCNN) network that implements a novel blurring-adjusting strategy to address the complex denoising problem via two stages. In stage 1, we build a blurring module (BM) to capture over-smoothed features from noisy observations and generate the blurred image restoration, which is a less informative version of the clean image. Furthermore, a multi-level concatenating module (CM) and an adjusting module (AM) are then designed to recover more detailed information in stage 2. These two modules are jointly designed to restore a properly-smoothed image from the over-smoothed blurred image and the given under-smoothed noisy image. Comparing to the traditional denoising process, the proposed blurring-adjusting strategy produces a precise denoised image more efficiently by converting the unidirectional denoising process into a bidirectional denoising process. To our knowledge, this is the first study that utilizes the over-smoothed image to address the denoising problem. Extensive experiments demonstrate the superiority of our BFCNN with more accurate reconstruction quality and achieve competitive quantitative results among current CNN-based methods. This research will release the code upon acceptance.

Learning Blind Denoising Network for Noisy Image Deblurring.

Spatial-Adaptive Network for Single Image Denoising

Multinoise-type Blind Denoising Using a Single Uniform Deep Convolutional Neural Network.

Bidirectional Image Denoising with Blurred Image Feature

Learning a Non-blind Deblurring Network for Night Blurry Images

Nonblind Image Deblurring Via Deep Learning in Complex Field

Image Denoising Based On A Cnn Model

Toward convolutional blind denoising of real photographs

DCANet: Dual Convolutional Neural Network with Attention for Image Blind Denoising

Dual convolutional neural network with attention for image blind denoising

Attention-guided CNN for image denoising

Image denoising using deep CNN with batch renormalization

Image Blind Denoising Using Dual Convolutional Neural Network with Skip Connection

Image Denoising Based On Edge Enhancement And Convolutional Neural Network

A Concatenated Residual Network for Image Deblurring.

Blind Image Denoising Via Dynamic Dual Learning

A Robust Deformed Convolutional Neural Network (CNN) for Image Denoising.

Deep Universal Blind Image Denoising

Unpaired Learning of Deep Image Denoising

CNB Net: A Two-Stage Approach for Effective Image Deblurring

An Adaptive CNN for Image Denoising