Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Solving General Noisy Inverse Problem via Posterior Sampling: A Policy Gradient Viewpoint

About

Solving image inverse problems (e.g., super-resolution and inpainting) requires generating a high fidelity image that matches the given input (the low-resolution image or the masked image). By using the input image as guidance, we can leverage a pretrained diffusion generative model to solve a wide range of image inverse tasks without task specific model fine-tuning. To precisely estimate the guidance score function of the input image, we propose Diffusion Policy Gradient (DPG), a tractable computation method by viewing the intermediate noisy images as policies and the target image as the states selected by the policy. Experiments show that our method is robust to both Gaussian and Poisson noise degradation on multiple linear and non-linear inverse tasks, resulting into a higher image restoration quality on FFHQ, ImageNet and LSUN datasets.

Haoyue Tang, Tian Xie, Aosong Feng, Hanyu Wang, Chenyang Zhang, Yang Bai• 2024

Related benchmarks

TaskDatasetResultRank
Posterior SamplingLinear Gaussian n=400
SWD6.111
24
Posterior SamplingLinear Gaussian n=80
SWD6.786
24
Posterior SamplingLinear Gaussian n=2
SWD3.955
24
Phase RetrievalFFHQ 256x256 (test)
LPIPS0.663
16
Inverse ProblemInverseBench Navier-Stokes (sigma_noise=0) (test)
Relative L2 Error0.325
13
Inverse ProblemInverseBench Navier-Stokes (sigma_noise=1.0) (test)
Relative L2 Error0.408
13
Inverse ProblemInverseBench Navier-Stokes (sigma_noise=2.0) (test)
Relative L2 Error0.466
13
Fluid Data Assimilationvorticity fields 128x128 (test)
Relative L2 Error0.491
12
Image DeblurringFFHQ Gaussian deblurring 100 images (val)
PSNR24.11
9
Image Super-resolutionFFHQ 4x super-resolution 100 images (val)
PSNR24.81
9
Showing 10 of 16 rows

Other info

Follow for update