AI Paper Atlas

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

Steering Guidance for Personalized Text-to-Image Diffusion Models

Sunghyun Park, Seokeon Choi, Hyoungwoo Park, Sungrack Yun

ICCV 2025 · Image generation › Text-to-image diffusion · approach: Compositional text-to-image guidance · 2 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 pre-trained model with null prompt as the weak model instead of the fine-tuned unconditional model (personalization guidance vs CFG)DB-LoRA on SANA · ViCoDINO: 0.4819 -> 0.5366Improved
use personalization guidance instead of CFGDB-LoRA on SD 2.1 · ViCoCLIP-T: 0.3323 -> 0.3254Hurt
lower the weak-model weight interpolation omega from 1.0 toward 0.6 (ablation)Personalization guidance on SD 1.5 DB-LoRA · ViCoDINO: 0.4701 -> 0.5024Improved
increase omega toward 1.0 (more fine-tuned weak model) (ablation)Personalization guidance on SD 1.5 DB-LoRA · ViCoCLIP-T: 0.3260 -> 0.3345Improved
increase omega on SANA DB-LoRA (ablation)Personalization guidance on SANA · ViCoDINO: 0.5366 -> 0.4819Hurt
add AutoGuidance on top of CFGDB-LoRA SD 1.5 · ViCoDINO: 0.4701 -> 0.4831Improved
increase CFG guidance scale lambda vs personalization guidance (ablation)SANA DB-LoRA · ViCoDINOMixed
combine personalization guidance with CFG++DB-LoRA SD 2.1 · ViCoDINO: 0.5105 -> 0.5385Improved
apply personalization guidance to Diffusion-DPODiffusion-DPO SD 1.5 · PartiPromptsPickScore: 21.53 -> 21.58Improved
apply personalization guidance to PairCustom style LoRAPairCustom · StyleAlign promptsDreamSim: 0.7993 -> 0.7344Improved
personalization guidance versus CFG in DB-LoRA training steps (ablation)SANA DB-LoRA · ViCosubject and text fidelityImproved
personalization guidance adds no overhead versus AG/SAGSD 2.1 DB-LoRA+TI · Inference costinference time: 21.2 secs -> 14.5 secsImproved

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