《计算机应用》唯一官方网站 ›› 2025, Vol. 45 ›› Issue (10): 3187-3194.DOI: 10.11772/j.issn.1001-9081.2024101445

• 人工智能 • 上一篇    

基于动态图卷积网络的电梯乘客异常行为数据增强方法

况世雄, 姚俊波, 陆佳炜, 王琪冰, 肖刚()   

  1. 中国计量大学 机电工程学院,杭州 310018
  • 收稿日期:2024-10-14 修回日期:2024-12-30 发布日期:2025-01-06 出版日期:2025-10-10
  • 通讯作者: 肖刚
  • 作者简介:况世雄(2000—),男,湖北武汉人,硕士研究生,主要研究方向:数据增强、模式识别
    姚俊波(2000—),男,浙江宁波人,硕士研究生,主要研究方向:信号处理、故障诊断
    陆佳炜(1981—),男,浙江湖州人,副教授,博士研究生,主要研究方向:知识图谱、深度学习
    王琪冰(1978—),男,浙江绍兴人,教授,博士,主要研究方向:数字化设计与制造、特种设备智能化
    肖刚(1965—),男,浙江上虞人,教授,博士,CCF会员,主要研究方向:故障诊断、智能制造。 Email:xg@zjut.edu.cn
  • 基金资助:
    浙江省“尖兵”研发攻关计划项目(2023C01022);浙江省“领雁”研发攻关计划项目(2023C01215);湖州市重点研发科技计划项目(2022ZD2019)

Data augmentation method for abnormal elevator passenger behaviors based on dynamic graph convolutional network

Shixiong KUANG, Junbo YAO, Jiawei LU, Qibing WANG, Gang XIAO()   

  1. College of Mechanical and Electrical Engineering,China Jiliang University,Hangzhou Zhejiang 310018,China
  • Received:2024-10-14 Revised:2024-12-30 Online:2025-01-06 Published:2025-10-10
  • Contact: Gang XIAO
  • About author:KUANG Shixiong, born in 2000, M. S. candidate. His research interests include data augmentation, pattern recognition.
    YAO Junbo, born in 2000, M. S. candidate. His research interests include signal processing, fault diagnosis.
    LU Jiawei, born in 1981, Ph. D. candidate, associate professor. His research interests include knowledge graph, deep learning.
    WANG Qibing, born in 1978, Ph. D., professor. His research interests include digital design and manufacturing, intelligence of special equipment.
    XIAO Gang, born in 1965, Ph. D., professor. His research interests include fault diagnosis, intelligent manufacturing.
  • Supported by:
    Zhejiang Province “Vanguard” Research and Development Breakthrough Program(2023C01022);Zhejiang Province “Leading Sparrow” Research and Development Breakthrough Program(2023C01215);Huzhou City Key Research and Development Science and Technology Program(2022ZD2019)

摘要:

目前,因缺乏足够多样化的异常行为数据,电梯乘客异常行为的识别方法存在准确率不高和泛化性能较差的问题。为解决这个问题,提出一种基于动态图卷积网络的行为数据增强方法(DGCN-BA)。首先,构建一种动态图卷积网络,用于捕捉电梯乘客行为中不同人体关节之间的空间关系和运动关联性;其次,利用这些特征进行姿势增强,获取更丰富和合理的姿势序列;最后,利用姿势序列在虚拟电梯场景中构建人物动作,生成大量电梯乘客异常行为视频数据。在公开数据集Human3.6M、3DHP和MuPoTS-3D,以及自建数据集上验证DGCN-BA的有效性。实验结果表明,相较于JMDA(Joint Mixing Data Augmentation)、DDPMs(Denoising Diffusion Probabilistic Models)数据增强方法,DGCN-BA在Human3.6M数据集上的平均每个关节位置误差(MPJPE)分别降低了2.9 mm和1.5 mm。可见,DGCN-BA能够更有效完成姿势估计任务,生成合理多样的异常行为数据,从而明显改善基于视频的电梯乘客异常行为识别效果。

关键词: 图卷积网络, 数据增强, 行为识别, 姿势估计, 电梯乘客

Abstract:

The problems of low accuracy and poor generalization performance in recognizing abnormal behaviors of elevator passengers are due to the lack of sufficient diverse abnormal behavior data. To address these issues, a Dynamic Graph Convolutional Network-based Behavior data Augmentation (DGCN-BA) method was proposed. Firstly, a dynamic graph convolutional network was constructed to capture spatial relationships and motion correlations among different human joints in the behaviors of elevator passengers. Secondly, these features were utilized to enhance pose data, thereby generating richer and more reasonable pose sequences. Finally, the pose sequences were used to construct human actions in a virtual elevator scene, and lot of abnormal behavior video data for elevator passengers were generated. To validate the effectiveness of DGCN-BA, experiments were conducted on public datasets Human3.6M, 3DHP, MuPoTS-3D, and a self-constructed dataset. Experimental results show that compared to data augmentation methods JMDA (Joint Mixing Data Augmentation) and DDPMs (Denoising Diffusion Probabilistic Models), DGCN-BA reduces the Mean Per Joint Position Error (MPJPE) on the Human3.6M dataset by 2.9 mm and 1.5 mm, respectively. It can be seen that DGCN-BA can complete pose estimation tasks more effectively, generates diverse and reasonable abnormal behavior data, and improves the recognition effect of video-based elevator passenger abnormal behaviors significantly.

Key words: graph convolutional network, data augmentation, behavior recognition, pose estimation, elevator passenger

中图分类号: