Stochastic Image Denoising by Sampling from the Posterior Distribution
This addresses the blurriness issue in image denoising for applications requiring visual quality, though it is incremental as it builds on existing MMSE denoisers.
The paper tackles the problem of blurry outputs in image denoising at high noise levels by proposing a stochastic method that samples from the posterior distribution using Langevin dynamics and repeated MMSE denoiser applications, producing high perceptual quality results with small MSE and enabling multiple legitimate outputs for a single input.
Image denoising is a well-known and well studied problem, commonly targeting a minimization of the mean squared error (MSE) between the outcome and the original image. Unfortunately, especially for severe noise levels, such Minimum MSE (MMSE) solutions may lead to blurry output images. In this work we propose a novel stochastic denoising approach that produces viable and high perceptual quality results, while maintaining a small MSE. Our method employs Langevin dynamics that relies on a repeated application of any given MMSE denoiser, obtaining the reconstructed image by effectively sampling from the posterior distribution. Due to its stochasticity, the proposed algorithm can produce a variety of high-quality outputs for a given noisy input, all shown to be legitimate denoising results. In addition, we present an extension of our algorithm for handling the inpainting problem, recovering missing pixels while removing noise from partially given data.