RAG / 检索增强生成

ImageAuditor: Membership Inference Attack against Image-based Retrieval-Augmented Generation

ImageAuditor: 针对基于图像的检索增强生成系统的成员推理攻击

Jinghuai Zhang, Pengyue Yu, Zhexiao Lin, Kunlin Cai, Fnu Suya, Yuan Tian

专题命中 多模态RAG ：retrieval-augmented generation(title,abstract)；RAG(abstract,abstract_cn)

AI总结 提出首个针对基于图像的检索增强生成（IRAG）的成员推理攻击方法ImageAuditor，通过奖励引导策略优化解决跨模态检索和判别信号提取两大挑战，在仅需四次查询时AUROC超过80%。

CourseTimeQA: A Lecture-Video Benchmark and a Latency-Constrained Cross-Modal Fusion Method for Timestamped QA

CourseTimeQA: 一个讲座视频基准和一种用于时间戳问答的延迟约束跨模态融合方法

Vsevolod Kovalev, Parteek Kumar

专题命中 多模态RAG ：RAG(summary_cn,abstract)；retriever(abstract)；分类 cs.CL

AI总结 针对教育讲座视频中的时间戳问答任务，在单GPU延迟/内存预算下，提出了CourseTimeQA基准和一种轻量级延迟约束跨模态检索器CrossFusion-RAG，通过冻结编码器、浅层查询无关交叉注意力和时间一致性正则化，在nDCG@10和MRR上分别提升0.10和0.08，中位端到端延迟约1.55秒。

Comments This paper is being withdrawn because an error in our measurement procedure produced incorrect values in our reported retrieval results (Tables I, II, V, and VI, and the corresponding headline figures in the Abstract). Several of our empirical claims depend on these measurements and therefore do not hold as stated