PRCC

该数据集名为PRCC，包含了221个身份标识，涵盖三个不同的摄像头视角，旨在针对衣物变化下的人物重新识别任务。数据集包含两种查询方式：同衣物查询和跨衣物查询，并且每种查询方式都有具体的数量统计。规模上，数据集包含221个身份，17896张训练图像，3384张图库图像，以及71个身份用于查询。该数据集的任务是衣物变化下的人物重新识别。Named PRCC, this dataset contains 221 unique identities and covers three distinct camera viewpoints, which is specifically developed for the task of person re-identification under clothing variation. It supports two query modes: same-clothing query and cross-clothing query, with specific quantitative statistics provided for each mode. In terms of dataset scale, it consists of 17,896 training images, 3,384 gallery images, and reserves 71 identities for querying, with a total of 221 identities included overall. The core task of this dataset is person re-identification under clothing variation.

搜集汇总

背景与挑战

背景概述

PRCC数据集是一个针对跨服装变化人员重识别的大规模数据集，包含33698张图像和221个身份，特别设计以支持不同服装条件下的人员重识别研究。数据集仅限于学术使用，版权归Sun Yat-sen University所有。