GPR1200

该数据集是一个用于内容为基础的图像检索的通用基准测试集，它包含了来自不同领域的子集。其中包含了1200个类别，每个类别有10张图片，用于评估图像检索的性能。该数据集的任务是零样本图像到图像的检索。

This dataset is a general-purpose benchmark for content-based image retrieval. It encompasses subsets from various domains, totaling 1200 categories with 10 images per category, and is utilized to evaluate the performance of image retrieval systems. The core task of this dataset is zero-shot image-to-image retrieval.