analiza aiqbench
CURE introduces a data-efficient curriculum learning framework that significantly improves visual grounding and factual consistency in medical vision-language models without requiring additional training data. By dynamically prioritizing harder samples, the method reduces hallucinations by 18.6% and enhances spatial-textual alignment, offering a practical solution for reliable radiology report automation. This approach addresses critical reliability issues in existing models through multi-task fine-tuning on public datasets.
CURE: Curriculum-guided Multi-task Training for Reliable Anatomy Grounded Report Generation
arXiv:2601.15408v2 Announce Type: replace-cross Abstract: Medical vision-language models can automate the generation of radiology reports but struggle with accurate visual grounding and factual consistency. Existing models often misalign textual findings with visual evidence, leading to unreliable or weakly grounded predictions. We present CURE, an error-aware curriculum learning framework that improves grounding and report quality without any additional data. CURE fine-tunes a multimodal instructional model on phrase grounding, grounded report generation, and anatomy-grounded report generation using public datasets. The method dynamically adjusts sampling based on model performance, emphasizing harder samples to improve spatial and textual alignment. CURE improves grounding accuracy by +0.35 IoU, boosts report quality by +0.192 CXRFEScore, and reduces hallucinations by 18.6%. CURE is a data-efficient framework that enhances both grounding accuracy and report reliability. Code is available at https://github.com/PabloMessina/CURE and model weights at https://huggingface.co/pamessina/medgemma-4b-it-cure
Abstract:Medical vision-language models can automate the generation of radiology reports but struggle with accurate visual grounding and factual consistency. Existing models often misalign textual findings with visual evidence, leading to unreliable or weakly grounded predictions. We present CURE, an error-aware curriculum learning framework that improves grounding and report quality without any additional data. CURE fine-tunes a multimodal instructional model on phrase grounding, grounded report generation, and anatomy-grounded report generation using public datasets. The method dynamically adjusts sampling based on model performance, emphasizing harder samples to improve spatial and textual alignment. CURE improves grounding accuracy by +0.35 IoU, boosts report quality by +0.192 CXRFEScore, and reduces hallucinations by 18.6%. CURE is a data-efficient framework that enhances both grounding accuracy and report reliability. Code is available at this https URL and model weights at this https URL
Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG)
arXiv:2601.15408 [cs.CV]
(or arXiv:2601.15408v2 [cs.CV] for this version)
https://doi.org/10.48550/arXiv.2601.15408 Focus to learn more arXiv-issued DOI via DataCite
Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2026, pp. 36279-36289
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 255 cuvinte
Subiecte conexe
Digital States 2026: AI Moves From Experiment to Infrastructure
govtech.com · 24 sept. 2026
Granite 4.2 LLMs: How They're Built
Hugging Face · 25 aug. 2026
Links #5: 2026/07
LessWrong · 31 iul. 2026
Runway’s WorldPrompt and the Engineering of Real-Time Worlds
Latent Space · 25 sept. 2026