AI Paper Atlas

Research map / Image generation / Image restoration

Efficient Zero-shot Inpainting with Decoupled Diffusion Guidance

Badr MOUFAD, Yazid Janati, Navid Bagheri Shouraki, Alain Oliviero Durmus, Thomas Hirtz, Eric Moulines, Jimmy Olsson

ICLR 2026 · Image generation › Image restoration · approach: Inverse imaging with learned priors · 4 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 DING instead of FlowChefFlowChef · FFHQ 768 (Half)FID: 20.2 -> 9.6Improved
use DING instead of DiffPIRDiffPIR · FFHQ 768 (Half)cPSNR: 30.91 -> 31.03Hurt
compare DING to Blended on observed-region fidelityBlended · FFHQ 768 (Half)cPSNR: 31.32 -> 31.03Hurt
use DING instead of DiffPIRDiffPIR · DIV2K 768 (Half)FID: 41.1 -> 39.2Improved
use DING instead of DDNMDDNM · PIE-BenchFID: 61.4 -> 61.4No clear effect
use DING instead of BlendedBlended · PIE-BenchCLIP-ED: 23.15 -> 23.36Improved
use DING instead of fine-tuned SD3 inpaint (28 NFEs)SD3 Inpaint · FFHQ 512 (Half)cPSNR: 21.69 -> 31.40Improved
use DING instead of fine-tuned SD3 inpaintSD3 Inpaint · PIE-BenchFID: 68.7 -> 63.6Improved
give fine-tuned SD3 inpaint 56 NFEs instead of 28 (ablation)SD3 Inpaint · FFHQ 512 (Half)FID: 23.5 -> 23.7No clear effect
evaluate denoiser at proxy variable (2 NFEs/step) instead of reusing previous noise prediction (ablation)Delayed DING · DIV2K 768 (Half)FID: 43.9 -> 39.2Improved
evaluate denoiser at proxy variable instead of reusing previous noise prediction (ablation)Delayed DING · FFHQ 768 (Half)FID: 9.1 -> 9.6Mixed
use default fast-decaying eta schedule instead of rescaled DDPM (B) (ablation)DING · FFHQ 768 (Half)FID: 21.5 -> 9.6Improved

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