English

应用科学学报 ›› 2024, Vol. 42 ›› Issue (5): 884-892.doi: 10.3969/j.issn.0255-8297.2024.05.014

• 计算机科学与应用 • 上一篇    

跨通道交互注意力机制驱动的双流网络跨模态行人重识别

何磊1, 栗风永1, 秦川2   

  1. 1. 上海电力大学 计算机科学与技术学院, 上海 201306;
    2. 上海理工大学 光电信息与计算机工程学院, 上海 200093
  • 收稿日期:2022-11-22 发布日期:2024-09-29
  • 通信作者: 栗风永，副教授，研究方向为多媒体信息安全、人工智能安全、机器学习。E-mail:fyli@shiep.edu.cn E-mail:fyli@shiep.edu.cn
  • 基金资助:
    国家自然科学基金（No.U1936213）；上海市自然科学基金（No.20ZR1421600）资助

Cross-Modal Person Re-identification Driven by Cross-Channel Interactive Attention Mechanism in Dual-Stream Networks

HE Lei1, LI Fengyong1, QIN Chuan2   

  1. 1. College of Computer Science and Technology, Shanghai University of Electric Power, Shanghai 201306, China;
    2. School of Optical-Electrical and Computer Engineering, University of Shanghai for Science and Technology, Shanghai 200093, China
  • Received:2022-11-22 Published:2024-09-29

PDF

391

摘要： 现有的跨模态行人重识别方法不能同时兼顾模态间与模态内的目标行人差异，很难提升检索准确度。为解决这一问题，引入跨通道交互的注意力机制，增强行人特征的鲁棒提取能力，有效抑制冗余特征的提取并获得更具辨别力的特征表达。进一步，联合异质中心三元组损失、三元组损失和身份损失进行监督学习，有效结合了行人特征的跨模态类间差异和类内差异。实验证明了所提方法的有效性。与7个已有的经典方法相比，所提方法在两个标准数据集RegDB与SYSU-MM01上都取得了较好的性能效果。

关键词: 跨模态, 行人重识别, 卷积神经网络, 注意力机制

Abstract: Existing cross-modal person re-identification methods often fail to take into account the difference of target person between modes and within modes, making it difficult to further improve the retrieval accuracy. To solve this problem, this paper introduces the cross-channel interaction attention mechanism to enhance the robust extraction of person features, effectively suppresses the extraction of irrelevant features and achieves more discriminative feature expression. Furthermore, hetero-center triplet loss, triplet loss and identity loss are combined for supervised learning, effectively integrating the intermodal and intra-class differences in person features. Experimental results demonstrate the effectiveness of the proposed method, which outperforms seven existing methods on two standard datasets, RegDB and SYSU-MM01.

Key words: cross-modal, person re-identification, convolutional neural network, attention mechanism

中图分类号: