SURE Guided Posterior Sampling: Trajectory Correction for Diffusion-Based Inverse Problems
This work addresses a bottleneck in computational efficiency for inverse problems in fields like imaging, though it is incremental as it builds on existing diffusion model frameworks.
The paper tackles the problem of error accumulation in diffusion-based inverse problem solving, which typically requires hundreds of steps, by introducing SURE Guided Posterior Sampling (SGPS) to correct trajectory deviations, enabling high-quality reconstruction with fewer than 100 Neural Function Evaluations.
Diffusion models have emerged as powerful learned priors for solving inverse problems. However, current iterative solving approaches which alternate between diffusion sampling and data consistency steps typically require hundreds or thousands of steps to achieve high quality reconstruction due to accumulated errors. We address this challenge with SURE Guided Posterior Sampling (SGPS), a method that corrects sampling trajectory deviations using Stein's Unbiased Risk Estimate (SURE) gradient updates and PCA based noise estimation. By mitigating noise induced errors during the critical early and middle sampling stages, SGPS enables more accurate posterior sampling and reduces error accumulation. This allows our method to maintain high reconstruction quality with fewer than 100 Neural Function Evaluations (NFEs). Our extensive evaluation across diverse inverse problems demonstrates that SGPS consistently outperforms existing methods at low NFE counts.