Regularization by denoising: Bayesian model and Langevin-within-split Gibbs sampling

arXiv:2402.12292v1 Announce Type: cross
Abstract: This paper introduces a Bayesian framework for image inversion by deriving a probabilistic counterpart to the regularization-by-denoising (RED) paradigm. It additionally implements a Monte Carlo algorithm specifically tailored for sampling from the resulting posterior distribution, based on an asymptotically exact data augmentation (AXDA). The proposed algorithm is an approximate instance of split Gibbs sampling (SGS) which embeds one Langevin Monte Carlo step. The proposed method is applied to common imaging tasks such as deblurring, inpainting …

abstract algorithm arxiv augmentation bayesian cs.cv cs.lg data denoising distribution framework gibbs image paper paradigm posterior regularization sampling stat.ml type