HPatches

该数据集是一个全面的基准测试，它评估了描述符在三个任务上的表现：补丁验证、图像匹配和补丁检索，这些任务根据几何扭曲进行了分类。子任务被进一步划分为简单、困难和挑战性三个级别。评估指标包括针对检索任务的平均精度（mAP）。研究的任务包括补丁验证、图像匹配和补丁检索。

This dataset is a comprehensive benchmark that evaluates the performance of descriptors across three tasks: patch verification, image matching, and patch retrieval, which are categorized based on geometric distortions. Subtasks are further divided into three difficulty levels: easy, hard, and challenging. Evaluation metrics include mean Average Precision (mAP) for the retrieval task. The tasks under study include patch verification, image matching, and patch retrieval.