Person re-identification by unsupervised video matching

详细信息    查看全文

• 作者：Xiaolong Map>ap>p> ; p>p>dp>p> ; p>p>goup000@163.comp>Author Vitae ; Xiatian Zhup>bp>p> ; p>p>xiatian.zhu@qmul.ac.ukp>Author Vitae ; Shaogang Gongp>bp>p> ; p>p>s.gong@qmul.ac.ukp>Author Vitae ; Xudong Xiep>ap>p> ; p>p>xdxie@mail.tsinghua.edu.cnp>Author Vitae ; Jianming Hup>ap>p> ; p>p>hujm@mail.tsinghua.edu.cnp>Author Vitae ; Kin-Man Lamp>cp>p> ; p>p>kin.man.lam@polyu.edu.hkp>Author Vitae ; Yisheng Zhongp>ap>p> ; p>p>zys-dau@mail.tsinghua.edu.cnp>Author Vitae
• 关键词：Person re-identification ; Action recognition ; Gait recognition ; Video matching ; Temporal sequence matching ; Spatio-temporal pyramids ; Time shift
• 刊名：Pattern Recognition
• 出版年：2017
• 出版时间：May 2017
• 年：2017
• 卷：65
• 期：Complete
• 页码：197-210
• 全文大小：1762 K
• 卷排序：65

文摘

We propose an unsupervised approach to person re-identification based on typical surveillance image-sequences. We present a new video representation particularly tailored for person ReID. Specifically, this representation is built up on existing action space-time features. We introduce an effective video matching algorithm, Time Shift Dynamic Time Warping (TS-DTW) and its Multi-Dimension variant MDTS-DTW, for data selective based sequence matching.