AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

DiffVP:Differential Visual Semantic Prompting for LLM-Based CT Report Generation

Yuhe Tian, Kun Zhang, Haoran Ma, Rui Yan, Yingtai Li, Rongsheng Wang, S Kevin Zhou

ECCV 2026 · Vision-language and multimodal › Multimodal LLMs · approach: Multimodal LLM benchmarks · 1 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 diagnostic anchor E (predicted disease prompts) (ablation)baseline · RadGenome-ChestCTCE F1: 0.159 -> 0.256Improved
add global delta operator (ablation)baseline with E · RadGenome-ChestCTCE F1: 0.256 -> 0.411Improved
add local delta operator (ablation)baseline with E · RadGenome-ChestCTCE F1: 0.256 -> 0.406Improved
combine global and local delta (ablation)DiffVP · RadGenome-ChestCTBLEU-4: 33.20 / 33.89 -> 34.28Improved
add local delta on top of global delta, without E (ablation)DiffVP without E · RadGenome-ChestCTCE F1: 0.188 -> 0.238 F1; ROUGE-L 42.50 -> 32.83Mixed
increase visual prefix length beyond 16 (ablation)DiffVP · RadGenome-ChestCTBLEU-4: 34.28 (p=16) -> 33.32 (p=24) -> 31.53 (p=32)Hurt
decrease prefix length to 4 (ablation)DiffVP · RadGenome-ChestCTBLEU-4: 34.28 -> 29.72Hurt
compute discrepancy in latent token space instead of pixel space (ablation)DiffVP · RadGenome-ChestCTBLEU-1 and CE-F1Improved
DiffVP vs Reg2RG on RadGenome-ChestCTDiffVP · RadGenome-ChestCTBLEU-4: 24.87 -> 34.28Improved
DiffVP vs Reg2RG on ROUGE-LDiffVP · RadGenome-ChestCTROUGE-L: 36.65 -> 32.32Hurt
DiffVP vs Reg2RG on clinical precisionDiffVP · RadGenome-ChestCTCE Precision: 0.423 -> 0.366Hurt
joint end-to-end training of classifier and generator (ablation)DiffVP · RadGenome-ChestCTreport qualityHurt

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