SCPNet: Spatial-Channel Parallelism Network for Joint Holistic and Partial Person Re-identification

被引：59

作者：

Fan, Xing ^{[1
,2
]}

Luo, Hao ^{[1
,2
]}

Zhang, Xuan ^{[2
]}

He, Lingxiao ^{[3
]}

Zhang, Chi ^{[2
]}

Jiang, Wei ^{[1
]}

机构：

[1] Zhejiang Univ, Hangzhou, Zhejiang, Peoples R China

[2] Megvii Inc Fac, Beijing, Peoples R China

[3] Chinese Acad Sci, Inst Automat, Beijing, Peoples R China

来源：

COMPUTER VISION - ACCV 2018, PT II | 2019年 / 11362卷

关键词：

Person re-identification; Deep learning; Spatial-channel parallelism;

D O I：

10.1007/978-3-030-20890-5_2

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

Holistic person re-identification (ReID) has received extensive study in the past few years and achieves impressive progress. However, persons are often occluded by obstacles or other persons in practical scenarios, which makes partial person re-identification non-trivial. In this paper, we propose a spatial-channel parallelism network (SCPNet) in which each channel in the ReID feature pays attention to a given spatial part of the body. The spatial-channel corresponding relationship supervises the network to learn discriminative feature for both holistic and partial person re-identification. The single model trained on four holistic ReID datasets achieves competitive accuracy on these four datasets, as well as outperforms the state-of-the-art methods on two partial ReID datasets without training.

引用

页码：19 / 34

页数：16