MANTA
收藏资源简介:
MANTA数据集由新南威尔士大学等机构创建,是一个大规模的多视角和视觉-文本异常检测数据集,专门用于微小物体。该数据集包含超过137,338张多视角图像,涵盖38个类别,跨越农业、医药、电子、机械和食品五个典型领域。数据集的创建过程包括从多个领域收集微小物体,使用五台高分辨率相机从不同角度捕捉图像,并进行详细的标注。MANTA数据集的应用领域广泛,旨在解决微小物体异常检测中的挑战,特别是在农业、医药和电子等行业中的应用。
The MANTA dataset, created by institutions including the University of New South Wales, is a large-scale multi-view and visual-text anomaly detection dataset specialized for tiny objects. It comprises over 137,338 multi-view images, spanning 38 categories across five typical domains: agriculture, medicine, electronics, machinery, and food. The development of the MANTA dataset entails collecting tiny objects from diverse domains, capturing images from distinct angles using five high-resolution cameras, and performing detailed annotations. The MANTA dataset boasts a broad spectrum of application scenarios, with the goal of addressing the challenges in anomaly detection for tiny objects, particularly in industries such as agriculture, medicine, and electronics.




