BTAD

该数据集名为BTAD，是一个面向工业领域的真实世界无监督异常检测数据集。它包含了2540张图像，这些图像被分为三个工业产品类别。在训练集中，我们使用正常图像进行训练，而在测试集中则同时包含了正常和异常的图像。整个数据集的规模为2540张图像，所面临的任务是异常检测。

This dataset, named BTAD, is a real-world unsupervised anomaly detection dataset tailored for industrial scenarios. It contains 2540 images divided into three industrial product categories. Only normal images are used for model training in the training set, while the test set includes both normal and anomalous images. The total scale of this dataset is 2540 images, and the task it targets is anomaly detection.

搜集汇总

背景与挑战

背景概述

BTAD是一个工业领域无监督异常检测数据集，包含2540张图像，分为三个类别，训练集为正常图像，测试集包含正常和异常图像。