MultiView Multi-Illumination Anomaly Detection (M2AD)
收藏资源简介:
M2AD数据集是一个大规模的视觉异常检测数据集,由华中科技大学智能制造装备与技术国家重点实验室和密歇根大学机器人学系联合创建。该数据集包含119,880张高分辨率图像,旨在测试视觉异常检测系统在复杂视图-光照相互作用下的鲁棒性。数据集通过12个同步视图和10种照明设置(总共120种配置)系统地捕获了999个样本,覆盖了10个类别。M2AD数据集可用于开发和应用视觉异常检测系统,尤其是在工业质量控制、医学成像等领域。
The M2AD dataset is a large-scale visual anomaly detection dataset jointly developed by the State Key Laboratory of Intelligent Manufacturing Equipment and Technology at Huazhong University of Science and Technology and the Department of Robotics at the University of Michigan. It contains 119,880 high-resolution images, and is designed to evaluate the robustness of visual anomaly detection systems under complex view-light interactions. Specifically, the dataset systematically captures 999 samples across 10 categories, with data acquired via 12 synchronized views and 10 lighting settings, resulting in a total of 120 experimental configurations. The M2AD dataset can be utilized for developing and deploying visual anomaly detection systems, especially in domains such as industrial quality control and medical imaging.
- 1Visual Anomaly Detection under Complex View-Illumination Interplay: A Large-Scale Benchmark华中科技大学智能制造装备与技术国家重点实验室, 密歇根大学机器人学系 · 2025年



