Think Twice Before You Act: Improving Inverse Problem Solving With MCMC

Yaxuan Zhu,Zehao Dou,Haoxin Zheng,Yasi Zhang,Ying Nian Wu,Ruiqi Gao
2024-09-13
Abstract:Recent studies demonstrate that diffusion models can serve as a strong prior for solving inverse problems. A prominent example is Diffusion Posterior Sampling (DPS), which approximates the posterior distribution of data given the measure using Tweedie's formula. Despite the merits of being versatile in solving various inverse problems without re-training, the performance of DPS is hindered by the fact that this posterior approximation can be inaccurate especially for high noise levels. Therefore, we propose \textbf{D}iffusion \textbf{P}osterior \textbf{MC}MC (\textbf{DPMC}), a novel inference algorithm based on Annealed MCMC to solve inverse problems with pretrained diffusion models. We define a series of intermediate distributions inspired by the approximated conditional distributions used by DPS. Through annealed MCMC sampling, we encourage the samples to follow each intermediate distribution more closely before moving to the next distribution at a lower noise level, and therefore reduce the accumulated error along the path. We test our algorithm in various inverse problems, including super resolution, Gaussian deblurring, motion deblurring, inpainting, and phase retrieval. Our algorithm outperforms DPS with less number of evaluations across nearly all tasks, and is competitive among existing approaches.
Machine Learning
What problem does this paper attempt to address?
The problem that this paper attempts to solve is how to improve the accuracy of posterior distribution approximation when using pre - trained diffusion models to solve inverse problems, especially under high - noise levels. Existing methods such as Diffusion Posterior Sampling (DPS) can solve multiple inverse problems without retraining the model, but their performance is affected by inaccurate posterior approximation, especially under high - noise conditions. Therefore, the paper proposes a new algorithm - Diffusion Posterior MCMC (DPMC), which is based on Annealed Markov Chain Monte Carlo (Annealed MCMC) to improve the solution of inverse problems. By defining a series of intermediate distributions and performing multiple sampling steps at each noise level, DPMC aims to reduce the cumulative error along the path, thereby improving the sample quality. Experimental results show that DPMC outperforms DPS in various inverse problems and is competitive compared to existing methods in most tasks.