Blind Universal Bayesian Image Denoising with Gaussian Noise Level Learning

TL;DR

This paper introduces a theoretically-grounded deep learning model for blind, universal Gaussian image denoising that generalizes well to unseen noise levels and improves performance on real-world images.

Contribution

It presents a novel fusion denoising network based on Gaussian prior assumptions, enabling effective blind and universal denoising without prior noise level knowledge.

Findings

Generalizes to unseen noise levels in synthetic experiments

Improves PSNR on real-world grayscale images across noise levels

Enhances state-of-the-art color image denoising performance

Abstract

Blind and universal image denoising consists of using a unique model that denoises images with any level of noise. It is especially practical as noise levels do not need to be known when the model is developed or at test time. We propose a theoretically-grounded blind and universal deep learning image denoiser for additive Gaussian noise removal. Our network is based on an optimal denoising solution, which we call fusion denoising. It is derived theoretically with a Gaussian image prior assumption. Synthetic experiments show our network's generalization strength to unseen additive noise levels. We also adapt the fusion denoising network architecture for image denoising on real images. Our approach improves real-world grayscale additive image denoising PSNR results for training noise levels and further on noise levels not seen during training. It also improves state-of-the-art color image denoising performance on every single noise level, by an average of 0.1dB, whether trained on or not.