ADAM

该数据集名为自动检测年龄相关性黄斑变性挑战（ADAM），其中包含282张用于训练的正常图像和118张用于测试的异常图像。每个异常图像都包含了五种不同病变的遮罩，包括干性黄斑变性、渗出、出血、疤痕和其他病变。原始的眼底图像被调整为784×784像素，以便进行异常检测。该数据集的规模为282张训练图像和118张测试图像，任务是对眼底图像进行异常检测。

This dataset is named the Automatic Detection of Age-Related Macular Degeneration Challenge (ADAM). It contains 282 normal images for training and 118 abnormal images for testing. Each abnormal image is paired with masks for five distinct lesions, including dry age-related macular degeneration, exudation, hemorrhage, scar, and other lesions. The original fundus images were resized to 784×784 pixels to facilitate anomaly detection. The total scale of this dataset is 282 training images and 118 test images, and the task is to conduct anomaly detection on fundus images.