NAB

该数据集包含了23,929张训练图像和24,633张测试图像，展示了555种不同鸟类的图片。此外，数据集还提供了一个5级深的类别层次结构。规模上，数据集分为555个类别，共计48,562张图像。其任务是对图像进行分类和检索。

This dataset comprises 23,929 training images and 24,633 test images, depicting 555 distinct bird species. Additionally, the dataset provides a 5-level deep category hierarchy. In terms of scale, the dataset is structured into 555 categories with a total of 48,562 images. The core tasks of this dataset are image classification and image retrieval.