AI Paper Atlas

Research map / Image generation / Fast diffusion sampling

Decoupled Residual Denoising Diffusion Models for Unified and Data Efficient Image-to-Image Translation

Ziyue Lin, Jiahe Hou, Hongyu Xia, Xinrui Xie, Feifei Wang, Yuyin Zhou, Wei Wang, Jiawei Liu, Liangqiong Qu

CVPR 2026 · Image generation › Fast diffusion sampling · approach: Denoising diffusion probabilistic models · 1 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use DRDD instead of the best baseline DFPIRDRDD · All-in-One-5 (average)SSIM: 0.912 -> 0.916Improved
use DRDD instead of DFPIRDRDD · All-in-One-5 (average)LPIPS: 0.081 -> 0.073Improved
use DRDD instead of DA-CLIP (best FID baseline)DRDD · All-in-One-5 (average)FID: 20.0 -> 18.3Improved
use DRDD instead of VLUNet on DerainingDRDD · All-in-One-5 DerainingSSIM: 0.981 -> 0.978Hurt
use DRDD instead of DA-CLIP on DehazingDRDD · All-in-One-5 DehazingSSIM: 0.959 -> 0.972Hurt
use DRDD instead of VLUNet on MNMDDRDD · MNMD (average)SSIM: 0.9274 -> 0.9338Improved
use DRDD instead of VLUNet on MNMDDRDD · MNMD (average)LPIPS: 0.0784 -> 0.0553Improved
decouple noise and residual diffusion in IR-SDE (ablation)De-IRSDE · CelebA-HQ inpaintingFID: 15.14 -> 15.10Improved
decouple noise and residual diffusion in IR-SDE (ablation)De-IRSDE · Rain100 derainingPSNR: 27.2 -> 28.1Improved
decouple noise and residual diffusion in IR-SDE (ablation)De-IRSDE · BSD400 denoisingSSIM: 0.833 -> 0.827No clear effect
decouple noise and residual diffusion in IR-SDE (ablation)De-IRSDE · BSD400 denoisingFID: 33.29 -> 31.87Improved
set the injected noise intensity to 1.0 (ablation)DRDD · All-in-One-5SSIMImproved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy