Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

SURE Guided Posterior Sampling: Trajectory Correction for Diffusion-Based Inverse Problems

About

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.

Minwoo Kim, Hongki Lim• 2025

Related benchmarks

TaskDatasetResultRank
4x super-resolutionFFHQ 256x256
PSNR29.384
25
Gaussian deblurFFHQ 256x256
PSNR29.353
20
Inpainting (Box Mask)FFHQ 256x256
LPIPS0.133
13
Inpainting (Random Mask)FFHQ 256x256
LPIPS0.116
13
Nonlinear DeblurFFHQ
PSNR27.332
13
HDRFFHQ
PSNR24.872
13
Motion DeblurringFFHQ 256x256
LPIPS0.148
7
Phase RetrievalFFHQ
LPIPS0.268
7
Showing 8 of 8 rows

Other info

Follow for update