AI Paper Atlas

Research map / Vision-language and multimodal / CLIP and zero-shot

Ultrasound-CLIP: Semantic-Aware Contrastive Pre-training for Ultrasound Image-Text Understanding

Jiayun Jin, Haolong Chai, Xueying Huang, Xiaoqing Guo, Zengwei Zheng, Zhan Zhou, Junmei Wang, Xinyu Wang, Jie Liu, Binbin Zhou

CVPR 2026 · Vision-language and multimodal › CLIP and zero-shot · approach: CLIP training recipes and data curation · 5 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 UDAF-guided semantic loss and graph encoder to baseline CLIP training (ablation)Ultrasound-CLIP · US-365K retrieval (I2T)Recall@10: 0.2542 -> 0.3745Improved
add UDAF-guided semantic loss to graph-encoder-only model (ablation)Ultrasound-CLIP · US-365K retrieval (I2T)Recall@10: 0.2683 -> 0.3745Improved
add heterogeneous graph encoder to semantic-loss-only model (ablation)Ultrasound-CLIP · US-365K retrieval (I2T)Recall@10: 0.3444 -> 0.3745Improved
add both components versus neither (ablation)Ultrasound-CLIP · US-365K multi-task classificationaverage accuracy (%): 50.84 -> 59.61Improved
add graph encoder to semantic-loss-only model (ablation)Ultrasound-CLIP · US-365K multi-task classificationaverage accuracy (%): 49.87 -> 59.61Improved
add semantic loss to graph-encoder-only model (ablation)Ultrasound-CLIP · US-365K multi-task classificationaverage accuracy (%): 48.62 -> 59.61Improved
pretrain on US-365K with Ultrasound-CLIP instead of BiomedCLIPBiomedCLIP · US-365K multi-task classificationaverage accuracy (%): 33.81 -> 59.61Improved
pretrain with Ultrasound-CLIP versus BiomedCLIPBiomedCLIP · US-365K diagnosis classificationaccuracy (%): 39.40 -> 64.05Improved
Ultrasound-CLIP versus BiomedCLIP in linear probeBiomedCLIP · Downstream 4 datasets (linear probe)average accuracy (%): 74.14 -> 75.40Improved
Ultrasound-CLIP versus UniMed-CLIP in zero-shotUniMed-CLIP · Downstream 4 datasets (zero-shot)average accuracy (%): 49.44 -> 54.42Improved
Ultrasound-CLIP versus SigLIP on BreastMNIST zero-shotSigLIP · BreastMNIST (zero-shot)accuracy (%): 54.30 -> 39.16Hurt

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