清如的博客

Hetero-Center Loss for Cross-Modality Person Re-Identification目录Hetero-Center Loss for Cross-Modality Person Re-IdentificationAbstract1、Introduction2 、Our Approach2.1、问题描述2.2、异质中心损失（Hereo-Center Loss）2.3、双流局部特征网络3 、EXPERIMENTSAbstract跨模态行人重识别任务可以解决光照不佳环

Cross-Spectrum Dual-Subspace Pairing for RGB-infrared Cross-Modality Person Re-Identification目录Cross-Spectrum Dual-Subspace Pairing for RGB-infrared Cross-Modality Person Re-IdentificationAbstract1、Introduction2 、Our Approach2.1、Cross-spectrum Image Gener

Deep Learning for Person Re-identification:A Survey and Outlook写在前面：感谢叶茫博士对AGW的开源，AGW非常适合刚接触跨模态ReID的同学作为baseline，具有优异性能的同时，代码层面逻辑非常清晰，开源代码在文末附上链接。本文虽为综述类文章，但专栏的重点在于跨模态行人重识别。目录Deep Learning for Person Re-identification:A Survey and OutlookAbstractIntrod

Visible Thermal Person Re-Identification via Dual-Constrained Top-Ranking目录Visible Thermal Person Re-Identification via Dual-Constrained Top-RankingAbstractIntroductionProposed MethodExperimental ResultsAbstract提出了一种双向双重约束损失的双流网络。具有两方面优势：1）直接进行端到端特征学习，

RGB-Infrared Cross-Modality Person Re-Identification提示：这里可以添加系列文章的所有文章的目录，目录需要自己手动添加例如：第一章 Python 机器学习入门之pandas的使用提示：文章写完后，目录可以自动生成，如何生成可参考右边的帮助文档文章目录RGB-Infrared Cross-Modality Person Re-IdentificationAbstract一、Introduction二、SYSU-MM01三、跨模态网络比较四、结论