AI Paper Atlas

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

Training-free Diffusion Model Adaptation for Variable-Sized Text-to-Image Synthesis

Zhiyu Jin, Xuli Shen, Bin Li, Xiangyang Xue

NeurIPS 2023 · Image generation › Text-to-image diffusion · approach: Text-driven synthesis and semantic guidance · 6 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 proposed attention scaling factor instead of 1/sqrt(d)Stable Diffusion · 224*224 LAION-5BFID: 74.5742 -> 41.8925Improved
use proposed attention scaling factorStable Diffusion · 448*448FID: 19.9039 -> 19.4923Improved
use proposed attention scaling factorStable Diffusion · 768*768FID: 29.5974 -> 28.1372Improved
use proposed attention scaling factorLatent Diffusion · 128*128 LAION-400MFID: 65.2988 -> 57.5542Improved
use proposed attention scaling factorLatent Diffusion · 384*384FID: 20.5466 -> 20.3842Improved
use proposed attention scaling factorStable Diffusion · 224*224CLIP score: 0.2553 -> 0.2764Improved
use proposed attention scaling factorStable Diffusion · 448*448CLIP score: 0.3176 -> 0.3180No clear effect
use proposed attention scaling factorLatent Diffusion · 224*224CLIP score: 0.3094 -> 0.3096No clear effect
use proposed attention scaling factorStable Diffusion · 224*224 human evalconsistency score: 5.13 -> 6.77Improved
use proposed attention scaling factorStable Diffusion · 768*768 human evalconsistency score: 6.74 -> 7.56Improved
use proposed scaling factor instead of MultiDiffusionStable Diffusion · 224*224 FID-4KFID: 154.6925 -> 50.3515Improved
use proposed scaling factor instead of MultiDiffusionStable Diffusion · 768*768 FID-4KFID: 40.9270 -> 28.1372Improved

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