AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

Skin-R1: Clinical Knowledge-Guided Dermatological Diagnosis Using Vision-Language Models

Zehao Liu, Weijieying Ren, Jipeng ZHANG, Tianxiang Zhao, Jingxi Zhu, Xiaoting Li, Vasant Honavar

ECCV 2026 · Vision-language and multimodal › Multimodal LLMs · approach: Multimodal LLM benchmarks · 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 textbook-trajectory SFT before RL (ablation)Skin-R1 (Qwen2.5-VL-7B) · Disease diagnosis, ID avgaccuracy: 0.5843 -> 0.6385Improved
add textbook-trajectory SFT before RL (ablation)Skin-R1 (Qwen2.5-VL-7B) · Disease diagnosis, OOD avgaccuracy: 0.5674 -> 0.7171Improved
add textbook-trajectory SFT before RL (ablation)Skin-R1 (Qwen2.5-VL-7B) · Lesion condition classification, ID avgaccuracy: 0.4602 -> 0.6928Improved
SFT on SkinRationale trajectories (ablation)Qwen2.5-VL-7B · Disease diagnosis, ID avgaccuracy: 0.4367 -> 0.4743Improved
SFT on SkinRationale trajectories (ablation)Qwen2.5-VL-7B · Disease diagnosis, OOD avgaccuracy: 0.5027 -> 0.5153Mixed
SFT on SkinRationale trajectories (ablation)Qwen2.5-VL-7B · Lesion condition classification, ID avgaccuracy: 0.3698 -> 0.6271Improved
SFT on SkinRationale trajectories (ablation)Qwen2.5-VL-7B · Valid answer rate, ID diagnosisinvalid-answer rate: about 30% -> 2% invalidImproved
add granularity-wise hierarchical reward (vs standard binary reward) (ablation)Skin-R1 · Disease diagnosis, OOD avgaccuracy: 0.6386 -> 0.7171Improved
add granularity-wise hierarchical reward (vs standard binary reward) (ablation)Skin-R1 · Disease diagnosis, ID avgaccuracy: 0.6379 -> 0.6385No clear effect
add granularity-wise hierarchical reward (vs standard binary reward) (ablation)Skin-R1 · Lesion condition classification, ID avgaccuracy: 0.6658 -> 0.6928Improved
train RL longer (500 -> 1,500 steps) (ablation)Skin-R1 · Disease diagnosis, ID avgaccuracy: 0.5875 -> 0.6385Improved
train RL longer (500 -> 1,500 steps) (ablation)Skin-R1 · Disease diagnosis, OOD avgaccuracy: 0.6742 -> 0.7171Improved

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