AI Paper Atlas

Research map / Image generation / Fast diffusion sampling

Divide-and-Denoise: A Game-Theoretic Method for Fairly Composing Diffusion Models

Abhi Gupta, Polina Barabanshchikova, Vikas K Garg, Samuel Kaski, Tommi Jaakkola

ICML 2026 · Image generation › Fast diffusion sampling · approach: Diffusion theory and design space · 0 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
add proportional fairness constraints (ablation)Divide-and-Denoise, 2 models · Multi-object SD 2.0GenEval %images: 87.00% -> 88.50%Improved
add proportional fairness constraints (ablation)Divide-and-Denoise, 3 models · Multi-object SD 2.0GenEval %images: 51.75% -> 59.50%Improved
add proportional fairness constraints (ablation)Divide-and-Denoise, 3 models · Multi-object SD 2.0envy-free violation rate: 0.356 -> 0.017Improved
add proportional fairness constraints (ablation)Divide-and-Denoise, 2 models · Multi-object SD 2.0CLIP min: 21.59 -> 21.53No clear effect
learned dynamic allocation instead of fixed vertical stripes (ablation)Divide-and-Denoise vs MultiDiffusion, 3 models · Multi-object SD 2.0GenEval %images: 14.00% -> 59.50%Improved
learned allocation instead of fixed stripesDivide-and-Denoise vs MultiDiffusion · Descriptive conceptsGenEval %images: 27.00% -> 55.75%Improved
learned allocation instead of fixed stripesDivide-and-Denoise vs MultiDiffusion · Conflicting conceptsVQA joint: 0.819 -> 0.905Improved
team of single-concept models instead of one multi-concept modelDivide-and-Denoise vs Multi-Concept Diffusion · Multi-object SD 2.0 (3 models)GenEval %images: 14.75% -> 59.50%Improved
use attention-based utilities instead of score-based (ablation)Divide-and-Denoise · Stable Diffusionutility noiseImproved
add Divide-and-Denoise on DiTDiT class-conditioned models · ImageNet class pairsCLIP joint: 28.56 -> 29.03Improved
very large beta_t (regularization on allocation) (ablation)Divide-and-Denoise · design analysisallocationHurt

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