BriefGPT.xyz
Aug, 2017
分割与融合：一种用于人员再识别的重新排名方法
Divide and Fuse: A Re-ranking Approach for Person Re-identification
HTML
PDF
Rui Yu, Zhichao Zhou, Song Bai, Xiang Bai
TL;DR
该研究提出了一种用于人员再识别的“分治与利用”重排序框架，它可以利用从高维特征向量不同部分中提取的多样性进行融合排序。
Abstract
As
re-ranking
is a necessary procedure to boost
person re-identification
(re-ID) performance on large-scale datasets, the diversity of feature becomes crucial to person reID for its importance both on designing p
→