Hetero-Center loss for cross -modality person Re-identification

NEUROCOMPUTING

卷:386

Article

Yuanxin Zhu¹  Zhao Yang¹  Li Wang¹  Sai Zhao¹  Xiao Hu¹  Dapeng Tao² 

[1] Guangzhou Univ, Sch Mech & Elect Engn, Guangzhou, Peoples R China

[2] Yunnan Univ, Sch Informat Sci & Engn, Kunming, Yunnan, Peoples R China

关键词: Cross-modality person re-identification;    Hetero-Center loss;    Local feature;   

DOI  :  10.1016/j.neucom.2019.12.100

来源: Elsevier

【 摘 要 】

Cross-modality person re-identification is a challenging problem which retrieves a given pedestrian image in RGB modality among all the gallery images in infrared modality. The task can address the limitation of RGB-based person Re-ID in dark environments. Existing researches mainly focus on enlarging inter-class differences of feature to solve the problem. However, few studies investigate improving intraclass cross-modality similarity, which is important for this issue. In this paper, we propose a novel loss function, called Hetero-Center loss (HC loss) to reduce the intra-class cross-modality variations. Specifically, HC loss can supervise the network learning the cross-modality invariant information by constraining the intra-class center distance between two heterogenous modalities. With the joint supervision of Cross-Entropy (CE) loss and HC loss, the network is trained to achieve two vital objectives, inter-class discrepancy and intra-class cross-modality similarity as much as possible. Besides, we propose a simple and high- performance network architecture to learn local feature representations for cross-modality person re-identification, which can be a baseline for future research. Extensive experiments indicate the effectiveness of the proposed methods, which outperform state-of-the-art methods by a wide margin. (c) 2019 Elsevier B.V. All rights reserved.

【 授权许可】

Free