Toward Convolutional Blind Denoising of Real Photographs

CVPR 2019  ·  Shi Guo, Zifei Yan, Kai Zhang, WangMeng Zuo, Lei Zhang ·

While deep convolutional neural networks (CNNs) have achieved impressive success in image denoising with additive white Gaussian noise (AWGN), their performance remains limited on real-world noisy photographs. The main reason is that their learned models are easy to overfit on the simplified AWGN model which deviates severely from the complicated real-world noise model. In order to improve the generalization ability of deep CNN denoisers, we suggest training a convolutional blind denoising network (CBDNet) with more realistic noise model and real-world noisy-clean image pairs. On the one hand, both signal-dependent noise and in-camera signal processing pipeline is considered to synthesize realistic noisy images. On the other hand, real-world noisy photographs and their nearly noise-free counterparts are also included to train our CBDNet. To further provide an interactive strategy to rectify denoising result conveniently, a noise estimation subnetwork with asymmetric learning to suppress under-estimation of noise level is embedded into CBDNet. Extensive experimental results on three datasets of real-world noisy photographs clearly demonstrate the superior performance of CBDNet over state-of-the-arts in terms of quantitative metrics and visual quality. The code has been made available at

PDF Abstract CVPR 2019 PDF CVPR 2019 Abstract
Task Dataset Model Metric Name Metric Value Global Rank Uses Extra
Training Data
Result Benchmark
Denoising Darmstadt Noise Dataset CBDNet(Syn) PSNR 37.57 # 4
Color Image Denoising Darmstadt Noise Dataset CBDNet (Blind) PSNR (sRGB) 38.06 # 4
SSIM (sRGB) 0.9421 # 4
Image Denoising DND CBDNet PSNR (sRGB) 38.06 # 13
SSIM (sRGB) 0.942 # 13
Noise Estimation SIDD CBDNet PSNR Gap 8.30 # 5
Average KL Divergence 0.728 # 5
Image Denoising SIDD CBDNet PSNR (sRGB) 30.78 # 18
SSIM (sRGB) 0.801 # 20


No methods listed for this paper. Add relevant methods here