计算机科学

计算机科学 ›› 2025, Vol. 52 ›› Issue (11A): 241100155-8.doi: 10.11896/jsjkx.241100155

• 计算机图形学&多媒体 • 上一篇    下一篇

基于多目标追踪的视频无关人员自动识别

马一心1, 曾军皓2, 杨鑫岩3, 梁刚3   

  1. 1 中国航空工业集团公司成都飞机设计研究所 成都 610073
    2 中国电子科技集团公司第三十研究所 成都 610041
    3 四川大学网络空间安全学院 成都 610211
  • 出版日期:2025-11-15 发布日期:2025-11-10
  • 通讯作者: 梁刚(lianggang@scu.edu.cn)
  • 作者简介:myx416@gmail.com
  • 基金资助:
    国家自然科学基金联合项目(62162057);四川省自然科学基金(2025ZNSFSC0509);四川省科技厅重点研发项目(2023YFG0294);教育部地方项目(2023CDLZ-2)

Automatic Recognition of Irrelevant Individuals in Videos Based on Multi-object Tracking

MA Yixin1, ZENG Junhao2, YANG Xinyan3, LIANG Gang3   

  1. 1 AVIC Chengdu Aircraft Design & Research Institute,Chengdu 610073,China
    2 No.30 Research Institute of China Electronics Technology Group Corporation,Chengdu 610041,China
    3 School of Cyber Science and Engineering,Sichuan University,Chengdu 610211,China
  • Online:2025-11-15 Published:2025-11-10
  • Supported by:
    National Natural Science Foundation of China(62162057),National Natural Science Foundation of Sichuan(2025ZNSFSC0509),Sichuan Science and Technology Program(2023YFG0294) and Local Projects of the Ministry of Education(2023CDLZ-2).

PDF (PC)

11

摘要： 无关人员自动识别旨在检测并识别视频中的无关人员,以解决其隐私保护问题。现有的隐私保护方法通过提取高级视觉特征识别与主题无关的个人。然而,高级特征的提取会显著影响视频的处理效率,难以处理海量视频数据。同时,现有的单帧识别方法没有考虑目标的时序特征,导致准确率较低。因此,提出了一种自动识别算法以高效识别无关人员,引入了多目标追踪方法来判断人物与视频之间的相关性。该方法能够从个人运动轨迹的时间和空间两个维度提取5种轻量特征。此外,为了解决视频运动过程中遮挡和模糊带来的挑战,采用了基于观察的轨迹关联算法,旨在提高运动跟踪的准确性。在各种数据集上进行了实验验证,结果表明,所提出的方法在各种指标上相较于当前的先进方法表现出显著的提升,其中MOTA指标最高提高10.87个百分点,HOTA指标最高提高10.95个百分点,无关人员识别的准确率达到98.13%。

关键词: 无关人员识别, 隐私保护, 多目标追踪, 人脸检测, 深度学习

Abstract: Automatic identification of irrelevant individuals aims to detect and identify irrelevant persons in videos to solve their privacy protection issues.Existing privacy protection methods extract high-level visual features to identify individuals irrelevant to the subject.However,the extraction of high-level features significantly affects the processing efficiency of the video and makes it difficult to process massive video data.At the same time,the existing single-frame recognition method does not consider the temporal characteristics of the target,resulting in low accuracy.Therefore,this paper proposes an automatic recognition algorithm to efficiently identify irrelevant individuals,and introduces a multi-target tracking method to determine the correlation between people and videos.The method can extract five lightweight features from the time and space dimensions of the individual’s motion trajectory.In addition,in order to solve the challenges brought by occlusion and blur during video motion,an observation-based trajectory association algorithm is adopted to improve the accuracy of motion tracking.Extensive experiments conducted on various datasets demonstrate that the proposed method achieves significant improvements across multiple evaluation metrics compared to state-of-the-art approaches.Specifically,the MOTA metric shows a maximum improvement of 10.87 percentage points,the HOTA me-tric achieves a maximum increase of 10.95 percentage points,and the accuracy of irrelevant individuals recognition reaches 98.13%.

Key words: Irrelevant individuals recognition, Privacy protection, Multi-object tracking, Face detection, Deep learning

中图分类号: 

  • TP391
[1]FAKLARIS C,CAFARO F,HOOK S A,et al.Legal and Ethical Implications of Mobile Live-streaming Video Apps[C]//Proceedings of the 18th International Conference on Human-Computer Interaction with Mobile Devices and Services Adjunct.2016:722-729.
[2]YOUTUBE H.Blur Your Videos[EB/OL].[2024-12-30].ht-tps://support.google.com/youtube/answer/9057652?hl=en.
[3]INGRID.Redact Faces with Azure Media Analytics[EB/OL].[2024-12-30].https://learn.microsoft.com/zh-cn/previous-versions/media-services/previous/media-services-face-redaction.
[4]ZHOU J,PUN C M.Personal Privacy Protection Via Irrelevant Faces Tracking and Pixelation in Video Live Streaming[J].IEEE Transactions on Information Forensics and Security,2020,16:1088-1103.
[5]HASAN R,CRANDALL D,FRITZ M,et al.Automatically De-tecting Bystanders in Photos to Reduce Privacy Risks[C]//2020 IEEE Symposium on Security and Privacy(SP).IEEE,2020:318-335.
[6]DARLING D,LI A,LI Q.Automated Bystander Detection and Anonymization in Mobile Photography[C]//Security and Privacy in Communication Networks:16th EAI International Confe-rence,SecureComm 2020,Washington,DC,USA,Part I 16.Springer International Publishing,2020:402-424.
[7]ZHU M,ZHANG R,WANG H.Recognizing Irrelevant Faces in Short-form Videos Based on Feature Fusion and Active Learning[J].Neurocomputing,2022,501:694-704.
[8]MA W,WU X,ZHAO S,et al.FedSH:Towards Privacy-preser-ving Text-based Person Re-Identification[J].IEEE Transactions on Multimedia,2023,26:5065-5077.
[9]LIN J,DAI X,NAI K,et al.BRPPNet:Balanced Privacy Protection Network for Referring Personal Image Privacy Protection[J].Expert Systems with Applications,2023,233:120960.
[10]WU D,HAN W,WANG T,et al.Referring Multi-object Trac-king[C]//Proceedings of the IEEE/CVF Conference on Compu-ter Vision and Pattern Recognition.2023:14633-14642.
[11]BAI Y,ZHAO Z,GONG Y,et al.Artrackv2:Prompting Autore-gressive Tracker Where to Look and How to Describe[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2024:19048-19057.
[12]NGUYEN P,QUACH K G,KITANI K,et al.Type-to-track:Retrieve Any Object Via Prompt-based Tracking[C]//NeurIPS 2023.2023.
[13]ZADEH A,CHONG LIM Y,BALTRUSAITIS T,et al.Convolutional Experts Constrained Local Model for 3d Facial Landmark Detection[C]//Proceedings of the IEEE International Conference on Computer Vision Workshops.2017:2519-2528.
[14]GE Z,LIU S,WANG F,et al.Yolox:Exceeding Yolo Series in 2021[J].arXiv:2107.08430,2021.
[15]BEWLEY A,GE Z,OTT L,et al.Simple Online and Realtime Tracking[C]//2016 IEEE International Conference on Image Processing(ICIP).IEEE,2016:3464-3468.
[16]CAO J,PANG J,WENG X,et al.Observation-centric Sort:Rethinking Sort For Robust Multi-object Tracking[C]//Procee-dings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2023:9686-9696.
[17]DENDORFER P,REZATOFIGHI H,MILAN A,et al.CVPR19 Tracking and Detection Challenge:How Crowded Can It Get?[J].arXiv:1906.04567,2019.
[18]SHAO S,ZHAO Z,LI B,et al.Crowdhuman:A Benchmark for Detecting Human in a Crowd[J].arXiv:1805.00123,2018.
[19]SUNDARARAMAN R,DE ALMEIDA BRAGA C,MARCH-AND E,et al.Tracking Pedestrian Heads in Dense Crowd[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2021:3865-3875.
[20]MA Y X,PPIP Dataset[EB/OL].[2024-12-30].https://github.com/shulie/PPIP-dataset.
[21]LIN T Y,MAIRE M,BELONGIE S,et al.Microsoft Coco:Common Objects in Context[C]//Computer Vision－ECCV 2014:13th European Conference,Zurich,Switzerland,Part V 13.Springer International Publishing,2014:740-755.
[22]BERNARDIN K,STIEFELHAGEN R.Evaluating Multiple Object Tracking Performance:the Clear MotMetrics[J].EURASIP Journal on Image and Video Processing,2008,2008:1-10.
[23]RISTANI E,SOLERA F,ZOU R,et al.Performance Measures and a Dataset for Multi-target,Multi-camera Tracking[C]//European Conference on Computer Vision.Cham:Springer International Publishing,2016:17-35.
[24]LUITEN J,OSEP A,DENDORFER P,et al.Hota:A Higher Order Metric for Evaluating Multi-object Tracking[J].International Journal of Computer Vision,2021,129:548-578.
[25]ZHANG Y,WANG C,WANG X,et al.Fairmot:On the Fair-ness of Detection and Re-identification in Multiple Object Tracking[J].International Journal of Computer Vision,2021,129:3069-3087.
[26]DUAN K,BAI S,XIE L,et al.Centernet:Keypoint Triplets for Object Detection[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision.2019:6569-6578.
[27]LOHN-JARAMILLO J,RAY L,GRANGER R,et al.Cluster-tracker:An Efficiency-Focused Multiple Object Tracking Method[J/OL].https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4102945.
[28]VO X T,HOANG V D,NGUYEN D L,et al.Pedestrian Head Detection and Tracking via Global Vision Transformer[C]//International Workshop on Frontiers of Computer Vision.Cham:Springer International Publishing,2022:155-167.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
渝ICP备16013121号-4
地址：重庆市渝北区洪湖西路18号    邮编：401121  
电话：023-63500828（编辑部） 023-67039612（会议/专题） 023-67039623（财务/发票）
E-mail：jsjkx12@163.com
本系统由北京玛格泰克科技发展有限公司设计开发