Papers › Forward-only Diffusion Probabilistic Models

Forward-only Diffusion Probabilistic Models

22 May 2025arXiv:2505.16733archive 2025-07-28

Ziwei Luo, Fredrik K. Gustafsson, Jens Sjölund, Thomas B. Schön

This work presents a forward-only diffusion (FoD) approach for generative modelling. In contrast to traditional diffusion models that rely on a coupled forward-backward diffusion scheme, FoD directly learns data generation through a single forward diffusion process, yielding a simple yet efficient generative framework. The core of FoD is a state-dependent linear stochastic differential equation that involves a mean-reverting term in both the drift and diffusion functions. This mean-reversion property guarantees the convergence to clean data, naturally simulating a stochastic interpolation between source and target distributions. More importantly, FoD is analytically tractable and is trained using a simple stochastic flow matching objective, enabling a few-step non-Markov chain sampling during inference. The proposed FoD model, despite its simplicity, achieves competitive performance on various image-conditioned (e.g., image restoration) and unconditional generation tasks, demonstrating its effectiveness in generative modelling. Our code is available at https://github.com/Algolzw/FoD.

PaperPDFCode

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

Algolzw/FoD officialmentioned in papermentioned on GitHubpytorch report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Conditional Image GenerationImage DehazingImage GenerationImage RestorationLow-Light Image EnhancementSingle Image Deraining

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Image Generation CIFAR-10 FoD-ODE FID 5.01 #34 of 78 Archive leaderboard report
Low-Light Image Enhancement LOL FoD Average PSNR 21.61 #35 of 40 Archive leaderboard report
Low-Light Image Enhancement LOL FoD FID 41.31 #35 of 40 Archive leaderboard report
Low-Light Image Enhancement LOL FoD LPIPS 0.105 #35 of 40 Archive leaderboard report
Low-Light Image Enhancement LOL FoD SSIM 0.819 #35 of 40 Archive leaderboard report
Single Image Deraining Rain100H Fod w/ NMC FID 15.64 #3 of 19 Archive leaderboard report
Single Image Deraining Rain100H Fod w/ NMC LPIPS 0.041 #3 of 19 Archive leaderboard report
Single Image Deraining Rain100H Fod w/ NMC PSNR 33.63 #3 of 19 Archive leaderboard report
Single Image Deraining Rain100H Fod w/ NMC SSIM 0.941 #3 of 19 Archive leaderboard report
Single Image Deraining Rain100H FoD FID 14.10 #4 of 19 Archive leaderboard report
Single Image Deraining Rain100H FoD LPIPS 0.038 #4 of 19 Archive leaderboard report
Single Image Deraining Rain100H FoD PSNR 32.56 #4 of 19 Archive leaderboard report
Single Image Deraining Rain100H FoD SSIM 0.925 #4 of 19 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

Diffusion

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections