EAGLE：专家增强的注意力引导用于多模态大语言模型中的无调优工业异常检测
EAGLE: Expert-Augmented Attention Guidance for Tuning-Free Industrial Anomaly Detection in Multimodal Large Language Models
AI总结：
EAGLE通过整合专家异常检测器与冻结的MLLM，提出无调优框架，提升多模态大语言模型在工业异常检测中的准确率，且在MVTec-AD和VisA数据集上达到94.4%和88.1%的异常鉴别准确率。
摘要：
多模态大语言模型（MLLMs）可以通过语义描述和异常推理增强工业异常检测，但其在二分类检测精度上仍落后于专家异常检测器。现有方法通过微调MLLMs或训练桥接模块来对齐专家输出与MLLM输入，限制了跨backbone的灵活性。我们提出EAGLE，一个无调优框架，将专家异常检测器与冻结的MLLM整合。EAGLE包括阈值引导提示选择（TGPS），通过专家模型统计估计决策阈值并选择文本和视觉提示；以及置信度感知注意力锐化（CAAS），当专家置信度较低时将MLLM注意力转向视觉证据。除了提高精度外，我们分析了MLLM的注意力，并发现正确的异常预测与更强的关注于真实缺陷区域相关；EAGLE一致加强了这种对齐。在MVTec-AD和VisA上，EAGLE在不进行参数更新的情况下改进了五个MLLM backbone，分别达到94.4%和88.1%的异常鉴别准确率，并在不牺牲MLLM语义推理能力的情况下实现了与微调方法相竞争的性能。
英文摘要：
Multimodal large language models (MLLMs) can enrich industrial anomaly detection with semantic descriptions and anomaly reasoning, but they still lag specialist anomaly detectors in binary detection accuracy. Existing approaches address this gap by fine-tuning MLLMs or training bridging modules to align expert outputs with MLLM inputs, limiting flexibility across backbones. We propose EAGLE, a tuning-free framework that integrates expert anomaly detectors with frozen MLLMs. EAGLE consists of Threshold-Guided Prompt Selection (TGPS), which estimates a decision threshold from expert model statistics and selects textual and visual prompts, and Confidence-Aware Attention Sharpening (CAAS), which shifts MLLM attention toward visual evidence when expert confidence is low. Beyond improving accuracy, we analyze MLLM attention and find that correct anomaly predictions are associated with stronger focus on ground-truth defect regions; EAGLE consistently strengthens this alignment. On MVTec-AD and VisA, EAGLE improves five MLLM backbones without parameter updates, reaching up to 94.4\% and 88.1\% in anomaly discrimination accuracy, respectively, and achieving performance competitive with fine-tuning-based methods while largely preserving MLLM semantic reasoning ability.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
