UltraBench 2：面向超声视觉基础模型的稳健评测基准

AI 摘要

研究团队提出 UltraBench 2，一个面向超声图像分析的视觉基础模型评测基准，强调标准化、可复现性与易用性，并覆盖广泛的解剖部位与任务。基于该基准对现有超声视觉基础模型进行比较后发现，超声专用预训练模型在分类任务上仍保持领先，而最先进的通用模型在分割任务上已追平。该工作旨在解决超声基础模型评测碎片化、难以衡量进展的问题。

正文节选

proceedings UltraBench 2: Towards Robust Evaluation of Vision Foundation Models on Ultrasound Abstract Benchmarking is an increasingly critical part of research in machine learning and the domains where it is applied, including healthcare. Yet, despite the steady development of new ultrasound foundation models in recent years, the development of well-designed benchmarks to evaluate them has lagged behind. This deficiency has led to fragmented and inconsistent evaluations of competing models, mak

发布时间：2026-09-26 12:00

抓取时间：2026-09-25 12:31

来源机构：arXiv