AI Paper Atlas

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

StyleDoctor: Towards Specialist Reward Model for Style-centric Generation Tasks

Xilin He, Xiaole Xian, Xiangyu Yue, Muhammad Haris Khan

CVPR 2026 · Image generation › Text-to-image diffusion · approach: Preference optimization for diffusion

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 global style feature learning and unified preference learning (both) (ablation)StyleDoctor (Qwen2.5-VL-3B) · SPRData unseen testMM Und. Acc.: 66.54 -> 71.14Improved
add global style feature learning (contrastive stage) (ablation)StyleDoctor · SPRData unseen testMM Und. Acc.: 66.54 -> 69.63Improved
add unified preference learning (ablation)StyleDoctor · SPRData unseen testMM Und. Acc.: 66.54 -> 69.70Improved
add both components on style customization (ablation)StyleDoctor reward · Style customization CSDCSD: 0.69 -> 0.78Improved
use StyleDoctor reward in RL fine-tuning of OmniGen2OmniGen2 · Text-controlled style generationstyle accuracy (StyleDoctor/GPT-4o): 65.42/58.85 -> 69.58/67.60Improved
use StyleDoctor reward in RL fine-tuning of Flux-KontextFlux-Kontext · Instruction-guided style transferGPT-4o style consistency: 74.35 -> 78.54Improved
use StyleDoctor reward in RL fine-tuning of OmniStyleOmniStyle · Reference image-guided style transferCSD: 0.74 -> 0.78Improved
replace CSD regularizer with StyleDoctor regularizerB-LoRA · Style customization (SDXL)CSD: 0.75 -> 0.78Improved
use human-preference reward model HPSv2 instead of StyleDoctor as rewardB-LoRA style customization · Style generationstyle consistency: 0.78 -> 0.45Hurt
use HPSv3 instead of StyleDoctor as rewardB-LoRA style customization · Style generationstyle consistency: 0.78 -> 0.64Hurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 4 of 4 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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