MirrorCheck: Efficient Adversarial Defense for Vision-Language Models 文章

ArXiv CS.CV2026-05-26NEWSen作者: Samar Fares, Klea Ziu, Toluwani Aremu, Nikita Durasov, Martin Tak\'a\v{c}, Pascal Fua, Ivan Laptev, Karthik Nandakumar
大语言模型
查看原文 →
关系图谱

MirrorCheck: Efficient Adversarial Defense for Vision-Language Models · 相关技术

相关技术

ginODEPLAlanguage model多模态图像编码器target modelrsaadversarial modelVerseText-to-imageSynthetic Aperture RadarStraight-Through EstimatorSpatial Pivot-Aligned Coordinate-free Embedding (SPACE)SSESPASEMParts-of-Speech (POS) tagsMulti-Instance RetrievalVision-Language ModelsVLMHISForFFIEvaluation Driven DevelopmentEffort Metric AttentionDAPTCamouflaged object detectionARGANN