Built independently by an author, for readers. Read the story and support ChapterPal
ChapterPal Logo
Log inSign up

keyword

DnCNN models

DnCNN models, short for Denoising Convolutional Neural Network models, are deep learning architectures designed to remove noise and visual artifacts from digital images. Rather than directly predicting the clean image from a degraded input, a DnCNN model utilizes residual learning to predict the underlying noise, which is subsequently subtracted from the input image to recover the restored output. These feed-forward networks integrate deep convolutional layers with batch normalization and rectified linear units to stabilize and accelerate training while boosting restoration performance. In addition to handling Gaussian noise across both known and blind noise levels, DnCNN models serve as versatile baselines for various image restoration tasks, including JPEG compression deblocking and single-image super-resolution.

1 item

Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising

Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising

Kai Zhang, Wangmeng Zuo, Yunjin Chen, Deyu Meng, Lei Zhang

OrganizationsGraz University of TechnologyHarbin Institute of TechnologyHong Kong Polytechnic UniversityXi'an Jiaotong University

Why you should read this

Introduces DnCNN, a deep convolutional neural network that combines residual learning and batch normalization to perform blind Gaussian image denoising with unknown noise levels and generalize across tasks like super-resolution and JPEG deblocking.

Discriminative model learning for image denoising has been recently attracting considerable attentions due to its favorable denoising performance. In this paper, we take one step forward by investigating the construction of feed-forward denoising convolutional neural networks (DnCNNs) to embrace the progress in very deep architecture, learning algorithm, and regularization method into image denoising. Specifically, residual learning and batch normalization are utilized to speed up the training process as well as boost the denoising performance. Different from the existing discriminative denoising models which usually train a specific model for additive white Gaussian noise (AWGN) at a certain noise level, our DnCNN model is able to handle Gaussian denoising with unknown noise level (i.e., blind Gaussian denoising). With the residual learning strategy, DnCNN implicitly removes the latent clean image in the hidden layers. This property motivates us to train a single DnCNN model to tackle with several general image denoising tasks such as Gaussian denoising, single image super-resolution and JPEG image deblocking. Our extensive experiments demonstrate that our DnCNN model can not only exhibit high effectiveness in several general image denoising tasks, but also be efficiently implemented by benefiting from GPU computing.

Added

2026-09-09