randall-lab/awa2
收藏官方服务：
资源简介：
AwA2（Animals with Attributes 2）数据集是一个包含37,322张图片的动物图像数据集，分为50个不同的动物类别。它最初是为了支持零样本学习、迁移学习和基于属性的分类任务而创建的。在这个仓库中，AwA2被用作一个标准的图像分类数据集，专注于图像和它们对应的动物类别标签，用于监督学习和传统深度学习模型的评估。
The Animals with Attributes 2 (AwA2) dataset is a collection of 37,322 images divided into 50 different animal categories. Originally designed for zero-shot learning, transfer learning, and attribute-based classification, it is now used in this repository as a standard image classification dataset focusing on image-label pairs for supervised learning and evaluation of traditional deep learning models.
提供机构：
randall-lab
