تخطي إلى التنقل الرئيسي تخطي إلى البحث تخطي إلى المحتوى الرئيسي

Graph-Based Temporal Attention Network for Anomaly Recognition in Internet of Things Video Surveillance

  • Waseem Ullah*
  • , Latif U. Khan
  • , Mohsen Guizani
  • , Chang Dong Wang
  • , Di Wu
  • *المؤلف المقابل لهذا العمل
  • Machine Learning Department
  • Mohamed Bin Zayed University of Artificial Intelligence
  • Department of Computer Science and Information Technology
  • Abu Dhabi University
  • Sun Yat-Sen University

نتاج البحث: Article › مراجعة النظراء

5 اقتباسات (Scopus)

ملخص

Accurate anomaly recognition in video surveillance is critical to ensuring public safety, infrastructure protection, and intelligent security operations. However, existing methods often fail to generalize across diverse and dynamic environments due to their limited capacity to model complex spatio-temporal patterns. In this article, we propose a novel graph-based temporal attention-memory network (G-TAMNet) that significantly advances video anomaly detection by capturing intricate temporal dependencies as well as spatial relationships. By integrating temporal self-attention into a graph convolutional network (GCN) framework, our model enhances the learning of salient and contextually relevant features that traditional approaches tend to overlook. Furthermore, to enable deployment in resource-constrained settings, such as Internet of Things (IoT)-based surveillance systems, we incorporate model quantization strategies that reduce computational overhead without compromising detection accuracy. Extensive evaluations of three widely used benchmarks - UCF-Crime, LAD-2000, and RWF-2000 - demonstrate the effectiveness of the proposed G-TAMNet model, achieving accuracy, precision, recall, and F1 -scores of 54.3%/61.1%/59.01%/61.1%, 79.3%/69.1%/70.0%/69.1%, and 94.0%/94.0%/94.2%/94.1%, respectively. These results reflect consistent performance gains of 3.3%, 9.9%, and 0.74% in accuracy over existing state-of-the-art methods on the corresponding datasets. Such improvements underscore the robustness, scalability, and practical viability of G-TAMNet for real-time anomaly detection in resource-constrained IoT surveillance systems. In addition, its reliable generalization across diverse environments highlights the potential of the model to advance intelligent security analytics and transform real-world applications in the domain of information forensics.

اللغة الأصليةEnglish
الصفحات (من إلى)18629-18643
عدد الصفحات15
دوريةIEEE Internet of Things Journal
مستوى الصوت13
رقم الإصدار9
المعرِّفات الرقمية للأشياء
حالة النشرPublished - 1 مايو 2026
منشور خارجيًانعم

أهداف الأمم المتحدة للتنمية المستدامة

يساهم هذا المخرج في تحقيق أهداف الأمم المتحدة للتنمية المستدامة التالية (SDGs)

  1. Industry innovation and infrastructure
    Industry innovation and infrastructure

بصمة

أدرس بدقة موضوعات البحث “Graph-Based Temporal Attention Network for Anomaly Recognition in Internet of Things Video Surveillance'. فهما يشكلان معًا بصمة فريدة.

قم بذكر هذا