Exact Posterior Score Estimation for Solving Linear Inverse Problems
Abbas Mammadov, Ozgur Kara, Kaan Oktay, Iskander Azangulov, Adil Kaan Akan, Hyungjin Chung, James Matthew Rehg, Yee Whye Teh
cs.LG, cs.CV, stat.ML
Submitted: 2026-06-15
License: http://creativecommons.org/licenses/by/4.0/
The gist: Diffusion and flow-based models learn powerful data priors by training a denoiser to reverse Gaussian corruption.
Terminology
Abstract
Diffusion and flow-based models learn powerful data priors by training a denoiser to reverse Gaussian corruption. To use this prior to solve a linear inverse problem, one needs to sample from the posterior, but the score that the prior provides is the unconditional score, not the posterior score. Existing methods either steer a fixed pretrained denoiser with approximate measurement-matching corrections, or train a conditional restoration model that abandons the denoising structure of the prior. We derive the exact posterior score in closed form for linear Gaussian inverse problems under general Gaussian interpolants, and show that posterior sampling reduces to a denoising problem at an operator-dependent shifted pivot under an anisotropic noise covariance. We turn this identity into Exact Posterior Score (EPS), a denoising training objective that preserves the input/output structure of standard pretraining and can therefore be trained from scratch or fine-tuned from a pretrained denoiser. At inference, EPS uses the same sampler as the underlying backbone, with no likelihood gradients or projections. We evaluate EPS on five linear inverse problems across FFHQ and ImageNet, where it outperforms training-free and training-based baselines on fidelity, perceptual, and distributional metrics, while using roughly an order of magnitude fewer denoiser evaluations than gradient-based posterior samplers.
Sources
- Deep Unsupervised Learning using Nonequilibrium Thermodynamics
- Score-Based Generative Modeling through Stochastic Differential Equations
- Flow Matching for Generative Modeling
- Flow Straight and Fast: Learning to Generate and Transfer Data with Rectified Flow
- Elucidating the Design Space of Diffusion-Based Generative Models
- Diffusion Posterior Sampling for General Noisy Inverse Problems
- Zero-Shot Image Restoration Using Denoising Diffusion Null-Space Model
- Manifold Preserving Guided Diffusion
- A Variational Perspective on Solving Inverse Problems with Diffusion Models
- A Survey on Diffusion Models for Inverse Problems
- Monte Carlo guided Diffusion for Bayesian linear inverse problems
- Conditional Image Generation with Score-Based Diffusion Models
- InvFusion: Bridging Supervised and Zero-shot Diffusion for Inverse Problems
- Inversion by Direct Iteration: An Alternative to Denoising Diffusion for Image Restoration
- I$^2$SB: Image-to-Image Schr\"odinger Bridge
- Amortized Posterior Sampling with Diffusion Prior Distillation
- Variational Flow Maps: Make Some Noise for One-Step Conditional Generation
- Distributional Diffusion Models with Scoring Rules
- FlowDPS: Flow-Driven Posterior Sampling for Inverse Problems
- GLASS Flows: Transition Sampling for Alignment of Flow and Diffusion Models
Related papers
- Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
- AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
- Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
- Convergence issues in Relational Concept Analysis based on AOC-posets
- Beliefs Beyond Posteriors: The Predictive Cost of ELBO Binding
- Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks