计算机工程与应用 ›› 2024, Vol. 60 ›› Issue (11): 156-164.DOI: 10.3778/j.issn.1002-8331.2304-0339

• 模式识别与人工智能 • 上一篇    下一篇

融合有效掩膜和局部增强的遮挡行人重识别

王小檬,梁凤梅   

  1. 太原理工大学 信息与计算机学院,山西 晋中 030600
  • 出版日期:2024-06-01 发布日期:2024-05-31

Effective Mask and Local Enhancement for Occluded Person Re-Identification

WANG Xiaomeng, LIANG Fengmei   

  1. College of Information and Computer, Taiyuan University of Technology, Jinzhong, Shanxi 030600, China
  • Online:2024-06-01 Published:2024-05-31

摘要: 在监控系统中行人经常会被各种障碍物遮挡,使得遮挡行人重识别仍然是一个长期存在的挑战。最近一些基于Transformer和外部语义线索的方法都改善了特征的表示和相关性能,但仍存在表示弱和语义线索不可靠等问题。为解决上述问题,提出了一种基于Transformer的新方法。引入了一种有效的掩膜生成方式,可靠的掩膜可以使模型不依赖外部语义线索并实现自动对齐。提出了一种基于平均注意力分数的序列重建模块,可以更有效地关注前景信息。提出了局部增强模块,获得了更鲁棒的特征表示。比较了所提方法和现有的各种方法在Occluded-Duke,Occluded-ReID,Partial-ReID,Market-1501数据集上的性能。Rank-1准确率分别达到了72.3%、84.8%、86.5%和95.6%,mAP精度分别为62.9%、83.2%、76.4%和89.9%,实验结果表明所提模型性能较其他先进网络有所提升。

关键词: 遮挡行人重识别, 原型掩膜, 特征注意力机制, 平均注意力分数, 局部增强, Transformer

Abstract: Human body is often occluded by a variety of obstacles in the monitoring system, so occluded person re-identi?cation is still a long-standing challenge. Recent methods based on Transformer and external semantic clues have improved feature representation and related performance, but there are still problems with weak representation and unreliable semantic clues. To solve the above problems, a novel method based on Transformer is proposed. Firstly, a more efficient way to generate masks is introduced. Reliable masks allow models to be independent of external semantic clues and to achieve automatic alignment. Secondly, a sequence reconstruction module based on average attention score is proposed, which can focus on foreground information more effectively. Thirdly, it proposes a local enhancement module to obtain more robust feature representation. Finally, performance of the propose method and various existing methods are compared on the Occluded-Duke, Occluded-ReID, Partial-ReID and Market-1501 datasets. The accuracy of Rank-1 of reaches 72.3%, 84.8%, 86.5% and 95.6%, respectively, the mAP accuracy are 62.9%, 83.2%, 76.4% and 89.9%. Experimental results demonstrate that the performance of the propose model is improved compared with other advanced networks.

Key words: occluded person re-identification, prototype mask, features attention mechanism, average attention score, local enhancement, Transformer