AwA2

该数据集名为AWA2，包含了37,322张图片，涉及50个动物类别（其中40个为已知类别，10个为未知类别），每个类别通过85个属性来描述。此外，该数据集专注于对动物物种进行粗粒度分类。规模上，数据集共有37,322张图片，其任务重点在于零样本学习。

This dataset, named AWA2, contains 37,322 images covering 50 animal categories, among which 40 are seen classes and 10 are unseen classes. Each category is described by 85 attributes. Additionally, this dataset focuses on coarse-grained classification of animal species, with its core task being zero-shot learning.