视觉大模型 / VLM

VisInject: Disruption != Injection -- A Dual-Dimension Evaluation of Universal Adversarial Attacks on Vision-Language Models

VisInject：破坏≠注入——对视觉-语言模型中通用对抗攻击的双维度评估

Pang Liu, Yingjie Lao

机构 * Department of Electrical and Computer Engineering, Tufts University（Tufts大学电气与计算机工程系） 纠错

专题命中 幻觉与鲁棒性 ：vision-language model(title)；multimodal large language model(abstract)；分类 cs.AI

AI总结 本文通过双维度评估揭示视觉-语言模型中通用对抗攻击的脆弱性，发现攻击成功与精确注入存在显著差异，揭示了模型在微小扰动下的表现。

Retrieval-Guided Generation for Safer Histopathology Image Captioning

基于检索的生成用于更安全的病理科图像描述生成

Md. Enamul Hoq, Wataru Uegami, Saghir Alfasly, Ghazal Alabtah, Sahar Rahimi Malakshan, Armita Kazemi, Alex T. Schmitgen, Fred Prior, H. R. Tizhoosh

机构 * Kimia Lab, Department of Artificial Intelligence \& Informatics, Mayo Clinic, Rochester, MN, USA ； Department of Biomedical Informatics, University of Arkansas for Medical Sciences, Little Rock, AR, USA ； Lane Department of Computer Science ； Electrical Engineering, West Virginia University, Morgantown, WV, USA ； Department of Computer Science ； Engineering, Princeton University, Princeton, NJ, USA ； Department of Computer Sciences, University of Wisconsin--Madison, Madison, WI, USA 纠错

专题命中 幻觉与鲁棒性 ：vision-language model(abstract)；分类 cs.CV、cs.AI

AI总结 本文提出检索引导生成方法，通过总结相似病例的专家文本生成描述，提升病理图像描述的准确性与可靠性，实验表明其在语义对齐和诊断一致性方面优于现有方法。