Image Blind Denoising with Generative Adversarial Network Based Noise Modeling

By Jingwen Chen, Jiawei Chen, Hongyang Chao, Ming-Hsuan Yang. 2018

568 citations recorded by the source corpus.

In this paper, we consider a typical image blind denoising problem, which is to remove unknown noise from noisy images. As we all know, discriminative learning based methods, such as DnCNN, can achieve state-of-the-art denoising results, but they are not applicable to this problem due to the lack of paired training data. To tackle the barrier, we propose a novel two-step framework. First, a Generative Adversarial Network (GAN) is trained to estimate the noise distribution over the input noisy images and to generate noise samples. Second, the noise patches sampled from the first step are utilized to construct a paired training dataset, which is used, in turn, to train a deep Convolutional Neural Network (CNN) for denoising. Extensive experiments have been done to demonstrate the superiority of our approach in image blind denoising.

Subfields: Computer Vision and Pattern Recognition, Media Technology

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

In Media Technology

Citations in the corpus

High-Fidelity Generative Image Compression (2020, background): “Beyond (un)conditional generation, adversarial losses led to advances in different image enhancement tasks, such as compression artifact removal [15], image de-noising [11], and image super-resolution [25].”
Image denoising using deep CNN with batch renormalization (2020, background): “The recently proposed generative adversarial network (GAN) is very popular for image denoising [6].”
Artificial intelligence in the creative industries: a review (2020, methodology): “lorization [82{84] [85{88] Super-resolution [89{103] [104,105] [90,103] Content Deblurring [106{112] [113] [110], Statistic model [114] Enhancement [115{117] BD [118,119] and Post Denoising [120{134] [128,135] [136,137] Filtering [138{142] Production Dehazing [143{147] [148,149] Turbulence removal [”
Deep Learning on Image Denoising: An overview (2019, methodology): “es Methods Applications Key words (remarks) Tao et al. (2019) [177] CNN Real noisy image denoising, low-light image enhancement CNN with ReLU, and RL for real noisy image denoising Chen et al. (2018) [28] CNN Real noisy image denoising, blind denoising GAN for real noisy image denoising Han et al. (”
Real-World Super-Resolution via Kernel Estimation and Noise Injection (2020, methodology): “Based on this observation and inspired by [6, 49], we de-sign a ﬁltering rule to collect patches with their variance in a certain range.”
Toward Convolutional Blind Denoising of Real Photographs (2018, unknown): “8646 GCBD [10] Blind sRGB 35.”
Self2Self With Dropout: Learning Self-Supervised Denoising From Single Image (2020, unknown): “[26, 31, 32, 8, 18, 13, 14].”
R2RNet: Low-light Image Enhancement via Real-low to Real-normal Network (2021, methodology): “Image denoising methods based on supervised learning, such as DnCNN-B [30], FFDNet [31], and CBDNet [32] utilized the Gaussian mixture model to perform denoising.”
Towards Real-World Blind Face Restoration with Generative Facial Prior (2021, unknown): “To achieve visually-pleasing results, generative adversarial network [19] is usually employed as loss supervisions to push the solutions closer to the natural manifold [43, 59, 66, 7, 15], while our work attempts to leverage the pretrained GANs as generative facial priors (GFP).”
Attention Guided Low-Light Image Enhancement with a Large Scale Low-Light Simulation Dataset (2019, methodology): “[11] propose a two-step framework which contains noise distribution estimation using GANs and denoising using CNNs.”
Deep Learning in Medical Imaging (2019, methodology): “Chen et al. 89 used GAN to denoising pipeline.”
DiffBIR: Towards Blind Image Restoration with Generative Diffusion Prior (2023, background): “GCBD [7] leverages generative adversarial networks (GAN) for noise modeling.”
Image Processing Using Multi-Code GAN Prior (2019, background): “These applications include image denoising [9, 25], image inpainting [43, 45], super-resolution [28, 41], image colorization [37, 20], style mixing [19, 10], semantic image manipulation [40, 29], etc .”