MulSen-AD
收藏资源简介:
MulSen-AD是由上海科技大学和密歇根大学安娜堡分校等机构联合创建的多传感器异常检测数据集,专门用于工业应用。该数据集整合了RGB图像、激光扫描的3D点云数据和锁相红外热成像数据,涵盖了15种工业产品,包含多种真实世界的异常情况。数据集的创建过程包括多传感器数据采集、处理和标注,确保了数据的多样性和准确性。MulSen-AD数据集的应用领域主要集中在工业质量检测,旨在通过多传感器融合技术提高异常检测的准确性和鲁棒性。
MulSen-AD is a multi-sensor anomaly detection dataset jointly developed by institutions including ShanghaiTech University and the University of Michigan, Ann Arbor, specifically tailored for industrial applications. It integrates RGB images, 3D point cloud data acquired via laser scanning, and lock-in infrared thermal imaging data, covering 15 types of industrial products and a variety of real-world anomaly scenarios. The dataset creation workflow encompasses multi-sensor data collection, processing and annotation, which ensures the diversity and accuracy of the collected data. The primary application field of the MulSen-AD dataset is industrial quality inspection, with the goal of improving the accuracy and robustness of anomaly detection through multi-sensor fusion technologies.




