2024年10月16日 · To bridge this gap, we present MMAD, the first-ever full-spectrum MLLMs benchmark in industrial Anomaly Detection. We defined seven key subtasks of MLLMs in industrial inspection and designed a novel pipeline to generate the MMAD dataset with 39,672 questions for 8,366 industrial images.

2024年10月12日 · To bridge this gap, we present MMAD, the first-ever full-spectrum MLLMs benchmark in industrial Anomaly Detection. We defined seven key subtasks of MLLMs in industrial inspection and designed a novel pipeline to generate the MMAD dataset with 39,672 questions for 8,366 industrial images.

为了回答这个问题，我们提出了MMAD（MLLM benchmark in industrial Anomaly Detection）——首个针对 工业异常检测 的多模态大模型基准测试。通过这项研究，我们发现了一些有趣的现象，探明了现有模型在工业异常检测中的表现，同时也揭示了这些模型在工业场景中的 …

2025年2月21日 · With MMAD, we have conducted a comprehensive, quantitative evaluation of various state-of-the-art (SOTA) MLLMs, including the GPT-4 series and Gemini 1.5 series (Reid et al., 2024), as well as open-source image models like InternVL2 (Chen et al., 2023) and LLaVA-NeXT (Liu et al., 2024a), and industry anomaly detection models like AnomalyGPT (Gu ...

2025年2月14日 · 最终构建的mmad数据集包含8,366张工业图像，涵盖38类产品和244种缺陷类型，生成39,672道多选问题，形成了工业领域最全面的mllm能力测评基准。 （左）MMAD数据集的数据信息，涵盖了7个关键子任务和38个代表性IAD类别。

With MMAD, we have conducted a comprehensive, quantitative evaluation of various state-of-the-art (SOTA) MLLMs, including the GPT-4 series and Gemini 1.5 series (Reid et al., 2024), as well as open-source image models like InternVL2 (Chen et …

2024年10月12日 · mmad数据集是由南方科技大学和腾讯优图实验室等机构联合创建的首个用于工业异常检测的多模态大语言模型综合基准。 该数据集包含39,672个多选题，基于8,366张工业图像，涵盖了38个工业产品类别和244种缺陷类型。

Security and privacy of electronic health records: Concerns and challenges. Egyptian Informatics Journal, 22 (2), 177-183. Innovations and Advances in Computing, Informatics, Systems Sciences …

2025年2月17日 · 为了回答这个问题，我们提出了 MMAD（MLLM benchmark in industrial Anomaly Detection）——首个针对工业异常检测的多模态大模型基准测试。 通过这项研究，我们发现了一些有趣的现象，探明了现有模型在工业异常检测中的表现，同时也揭示了这些模型在工业 …

2025年1月22日 · To bridge this gap, we present MMAD, a full-spectrum MLLM benchmark in industrial Anomaly Detection. We defined seven key subtasks of MLLMs in industrial inspection and designed a novel pipeline to generate the MMAD dataset with 39,672 questions for 8,366 industrial images.