Cloth-Changing Unreal Person (CCUP)
收藏arXiv2025-09-30 收录
下载链接：
https://github.com/yjzhao1019/CCUP
下载链接
链接失效反馈官方服务：
资源简介：
该数据集是一个包含6,000个身份、1,179,976张图片、100个摄像头以及每人平均26.5套服装的自标注CC-ReID数据集，专为训练在换衣人体重识别任务中的模型而设计。该数据集还被用于预训练如TransReID和FIRe^2等模型，相较于其他合成数据集，它在性能上有所提升。该数据集的规模属于大规模，其任务针对的是换衣人体重识别（Cc-Reid）。
This is a self-annotated CC-ReID dataset containing 6,000 identities, 1,179,976 images, 100 cameras, and an average of 26.5 outfits per identity. It is specifically designed for training models for cloth-changing person re-identification tasks. Moreover, this dataset has been used to pre-train models such as TransReID and FIRe^2, and demonstrates superior performance relative to other synthetic datasets. As a large-scale dataset, it is tailored for the cloth-changing person re-identification (Cc-ReID) task.
提供机构：
Research team at [Institute/University Name]
