Listening to the Noise: Blind Denoising with Gibbs Diffusion

arXiv:2402.19455v1 Announce Type: cross
Abstract: In recent years, denoising problems have become intertwined with the development of deep generative models. In particular, diffusion models are trained like denoisers, and the distribution they model coincide with denoising priors in the Bayesian picture. However, denoising through diffusion-based posterior sampling requires the noise level and covariance to be known, preventing blind denoising. We overcome this limitation by introducing Gibbs Diffusion (GDiff), a general methodology addressing posterior sampling of both the signal and the …

abstract arxiv astro-ph.co bayesian become blind cs.cv cs.lg deep generative models denoising development diffusion diffusion models distribution eess.sp generative generative models gibbs noise posterior sampling stat.ml through type