VC-Clothes

该数据集旨在针对不同相机视角下，穿着变化的人进行重识别测试。它特别适用于两组相机场景，包括同一服装和换装情景的测试。该数据集的任务是针对换装情况下的行人重识别。

This dataset is designed for person re-identification testing of individuals with clothing changes across different camera viewpoints. It is particularly suitable for two-camera scenario setups, covering both same-clothing and clothes-changing test cases. The core task of this dataset is person re-identification under clothes-changing conditions.

搜集汇总

背景与挑战

背景概述

VC-Clothes是一个行人重识别数据集，专门研究衣物更换对识别的影响，包含512个身份和19,060张图像，分为训练集和测试集。数据集还包含Real28子集，用于验证模型在真实场景中的表现。