AI Paper Atlas

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

DiffDoctor: Diagnosing Image Diffusion Models Before Treating

Yiyang Wang, Xi Chen, Xiaogang Xu, Sihui Ji, Yu Liu, Yujun Shen, Hengshuang Zhao

ICCV 2025 · Image generation › Text-to-image diffusion · approach: Inference-time T2I improvement · 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
add real photos as negatives to detector training (ablation)RichHF* detector · detector benchmarkMSE (percent): 1.601 -> 1.167Improved
add hard cases to detector training (ablation)RichHF* detector · detector benchmarkMSE (percent): 1.167 -> 0.504Improved
add 1M pseudo-labeled images to detector training (ablation)RichHF* detector · detector benchmarkMSE (percent): 0.504 -> 0.337Improved
add 1M pseudo-labeled images to detector training (ablation)RichHF* detector · detector benchmarkKL (false negative): 0.981 -> 1.004Hurt
use balanced best detector instead of naive RichHF* detector for treating (ablation)DiffDoctor · FLUX.1ImageReward: -0.884 -> 1.183Improved
use threshold mask instead of all pixels (ablation)DiffDoctor · FLUX.1ImageReward: 1.161 -> 1.183Improved
use threshold mask instead of max pixel (ablation)DiffDoctor · FLUX.1ImageReward: 1.159 -> 1.183Improved
add offline regularization at epoch 3 (ablation)DiffDoctor · FLUX.1ImageReward: 1.183 -> 1.167No clear effect
add offline regularization at epoch 10 (ablation)DiffDoctor · FLUX.1ImageReward: 0.904 -> 1.180Improved
add offline regularization at epoch 10 (ablation)DiffDoctor · FLUX.1CLIP-T: 34.952 -> 35.842Improved
treat FLUX.1 with DiffDoctorFLUX.1 · FLUX.1mean artifact frequency: 82.66% -> 22.00%Improved
treat SDXL with DiffDoctorSDXL · SDXLmean artifact frequency: 55.33% -> 27.50%Improved

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