AI Paper Atlas

Research map / Image generation / Text-to-image diffusion

Reusing Computation in Text-to-Image Diffusion for Efficient Generation of Image Sets

Dale Decatur, Thibault Groueix, Wang Yifan, Rana Hanocka, Vladimir Kim, Matheus Gadelha

ICCV 2025 · Image generation › Text-to-image diffusion · approach: Compositional text-to-image guidance · 3 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
share early diffusion steps across clustered prompts at fixed qualityKandinsky 2.2 · GenAI Benchcompute saved with similar VQA quality: 50.43% saved, 50.75% winImproved
share early diffusion steps across clustered prompts at fixed qualityKandinsky 2.2 · Style Variationscompute saved with similar VQA quality: 73.93% saved, 52.00% winImproved
use semantic clustering of prompt embeddings instead of random clustering (ablation)hierarchical diffusion · GenAI Benchcompute saved: 5.43% -> 50.43%Improved
use a model trained with text-to-image prior (Kandinsky) instead of SD 1.5 (ablation)hierarchical diffusion · Prompt Templatecompute saved: 28.25% -> 69.25%Improved
apply method to FLUX instead of Kandinsky (ablation)hierarchical diffusion · Prompt Templatecompute saved: 69.25% -> 3.61%Hurt
share computation under fixed budget instead of standard diffusionKandinsky 2.2 · Prompt Template VQA ScoreVQA Score at fixed compute budgetImproved
lower tau for earlier specialization (ablation)hierarchical diffusion · Prompt Templatecompute savings vs qualityMixed
use non-linear (log) mapping for phi instead of linear (ablation)hierarchical diffusion · text-to-image prompt setssavings vs qualityHurt
share computation on small homogeneous Style Variations sethierarchical diffusion · Style VariationsCLIP diversity (lower is more diverse): 0.8071 -> 0.8374Hurt
share computation on large prompt sets (500+)hierarchical diffusion · GenAI Bench / Prompt Template / AnimalsCLIP diversity: GenAI Bench 0.6082 -> 0.6027No clear effect

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