A Benchmark for Hallucination Detection in VLMs for Gastrointestinal Endoscopy
胃肠内窥镜中视觉语言模型幻觉检测的基准测试
机构 * University of Aberdeen(阿伯丁大学) ; Nepal Applied Mathematics and Informatics Institute for Research(尼泊尔应用数学与信息学研究所) ; West Virginia University(西弗吉尼亚大学)
专题命中 视觉问答 :VLM(summary_cn,abstract);LLaVA(abstract,abstract_cn);vision-language model(abstract);visual question answering(abstract)
AI总结 针对胃肠内窥镜领域,在Gut-VLM数据集上基准测试九种幻觉检测方法,发现白盒方法ReXTrust在所有五个视觉语言模型上AUC最高,平均领先19.5点。
Comments Accepted at the Medical Image Understanding and Analysis (MIUA) 2026 conference