Keywords
streaming anomaly detection
memory-regularized clustering
edge computing
video metric streams
online learning
operational anomaly detection
Abstract
Edge video analytics systems continuously generate metric streams from camera devices, object-detection modules, frame processing pipelines, network transmission, storage buffers, and inference accelerators. These metrics are non-stationary because of lighting variation, scene changes, camera movement, workload migration, and model update cycles. Abnormal conditions, such as camera tampering, inference service failure, dropped frames, and abnormal bandwidth consumption, may be hidden among repeated operational fluctuations. This study proposes a memory-regularized stream clustering method for anomaly detection in edge video analytics metrics. The method builds online clusters for normal metric patterns and uses a memory-regularization term to retain sensitivity to rare deviations over long-term monitoring. A multi-source metric embedding module combines frame rate, inference latency, object-count distribution, GPU utilization, buffer occupancy, packet loss, and video bitrate into streaming feature vectors. Experiments are conducted on an edge video analytics platform containing 2,460 cameras, 380 edge nodes, 96 inference services, and 44 monitoring indicators collected every 15 seconds over 64 days. The dataset includes 182 million metric records and 2,360 annotated abnormal episodes, including camera occlusion, frame-drop bursts, inference queue blockage, abnormal object-count drift, and edge-node network congestion. The proposed method shortens median detection delay from 12.4 minutes to 3.9 minutes compared with a sliding-window clustering baseline. False alerts are controlled at 1.6 cases per edge node per week. The system processes 41,000 metric records per second and maintains 6,780 active cluster states with 860 MB memory usage. Memory regularization preserves 1,540 low-frequency abnormal patterns that would otherwise be absorbed into normal operational clusters. The results demonstrate that memory-regularized stream clustering can improve anomaly detection for evolving edge video analytics data streams.
References
1. Fu, Y., Gui, H., Li, W., & Wang, Z. (2020, August). Virtual Material Modeling and Vibration Reduction Design of Electron Beam Imaging System. In 2020 IEEE International Conference on Advances in Electrical Engineering and Computer Applications (AEECA) (pp. 1063-1070). IEEE.
2. Puspa, S. N., & Chowdhury, M. (2025). GPU in the Blind Spot: Overlooked Security Risks in Transportation. arXiv preprint arXiv:2508.01995.
3. Liang, R., Feifan, F. N. U., Liang, Y., & Ye, Z. (2025). Emotion-Aware Interface Adaptation in Mobile Applications Based on Color Psychology and Multimodal User State Recognition. Frontiers in Artificial Intelligence Research, 2(1), 51-57.
4. Yahya, M. A., Moya, A. R., & Ventura, S. (2025). Deep learning for multivariate time series anomaly detection: An evaluation of reconstruction-based methods. Artificial Intelligence Review, 58(12), 400.
5. Chen, X., Xiao, H., Zeng, Z., Zhang, S., & Xiao, R. (2025). Fine-Grained Multivariate Time Series Anomaly Detection via Causal Inference. Knowledge-Based Systems, 114765.
6. Rasouli, N., Klein, C., & Elmroth, E. (2025). Resource management for mission-critical applications in edge computing: systematic review on recent research and open issues. ACM Computing Surveys, 58(3), 1-37.
7. Yang, M., Wang, Y., Shi, J., & Tong, L. (2025). Reinforcement Learning Based Multi-Stage Ad Sorting and Personalized Recommendation System Design.
8. Zhang, Z. (2026). A Study on Multimodal Product Understanding and Assembly Guidance Optimization in e-commerce for Complex Consumer Goods. Available at SSRN 6734559.
9. Yahya, M. A., Moya, A. R., & Ventura, S. (2025). Deep learning for multivariate time series anomaly detection: An evaluation of reconstruction-based methods. Artificial Intelligence Review, 58(12), 400.
10. Yin, J., Rao, H., & Huang, Y. (2026, March). Dynamic Modeling and Heterogeneity Analysis of Platform User Behavior Time Series. In 2026 International Conference on AI in Education Technology and Applications (AIETA) (pp. 30-33). IEEE.
11. Yang, J. (2026). Computational Analysis of How Digital Non-Clinical Communication Tools Influence Social Participation Among Older Adults in Community-Based Elderly Care. Available at SSRN 6682838.
12. Ortiz-Garces, I., Villegas-Ch, W., & Luján-Mora, S. (2025). Implementation of edge AI for early fault detection in IoT networks: Evaluation of performance and scalability in complex applications. Discover Internet of Things, 5(1), 108.
13. Urio-Larrea, A., Camargo, H., Lucca, G., Asmus, T., Marco-Detchart, C., Schick, L., ... & Dimuro, G. P. (2025). Data stream clustering: introducing recursively extendable aggregation functions for incremental cluster fusion processes. IEEE Transactions on Cybernetics, 55(3), 1421-1435.
14. Zhang, Z., Gao, Y., & Tong, Y. (2026). CausalML4CX: A Causal Inference and Machine Learning Framework for Customer Experience Optimization with Application in Banking.
15. Qi, C., & Qiao, X. (2026). Building and Operating a Large Scale Multi-Agent System: A Case Study from Industry. Available at SSRN 6795198.
16. Yin, J., Huang, Y., & Rao, H. (2026). A Study on the Dynamic Evolution of Learning Behavior and Outcome Prediction in Digital Educational Environments. Available at SSRN 6607139.
17. Ergenç, D., Memedi, A., Fischer, M., & Dressler, F. (2025). Resilience in edge computing: Challenges and concepts. Foundations and Trends in Networking, 14(4), 254-340.
18. Zhang, Z., Tong, Y., & Gao, Y. (2026). Causal Representation Learning for Explainable Early Warning in High-Stakes Decision Systems.
19. Yáñez-Sepúlveda, R., Vásquez-Bonilla, A., Olivares, R., Olivares, P., Zavala-Crichton, J. P., Hinojosa-Torres, C., ... & Garcia-Carrillo, E. (2025). Supervised machine learning algorithms for the classification of obesity levels using anthropometric indices derived from bioelectrical impedance analysis. Scientific Reports, 15(1).
20. Qi, C., & Qiao, X. (2026, May). Design Patterns for Multi-Agent Systems in Production. In 2026 International Conference on Intelligent Systems, Automation and Control (ISAC) (pp. 198-202). IEEE.
21. Su, D., & Shi, X. Optimizing Family–School Collaboration to Improve Educational and Social Outcomes for Children with Intellectual Disabilities in Inclusive Public Schools.
22. Kauser, S. Z., & Sable, A. H. (2026). Deep Learning Integration for Image Processing on Cloud Platforms: Enhancing Accuracy and Processing Speed through TensorFlow and PyTorch. International Journal of Aquatic Research and Environmental Studies, 6(S5), 730-741.
23. Zhang, Z. (2026). A Study on the Identification of Manipulative Design in Subscription and Payment Interfaces of Digital Consumer Platforms and Its Behavioral Effects. Available at SSRN 6734760.
24. Su, D., & Wu, Y. (2026). Multilevel Growth and Social Network Modeling for Functional Communication Enhancement in Students with Intellectual Disabilities.
25. Kirubavathi, G., Pulliyasseri, A., Rajesh, A., Ajayan, A., Alfarhood, S., Safran, M., ... & Shin, J. (2025). Enhancing IoT resilience at the edge: a resource-efficient framework for real-time anomaly detection in streaming data. Computer Modeling in Engineering & Sciences, 143(3), 3005.
26. Gao, G., Gao, R., Lu, C., Gao, R., & Kuang, Y. (2026, March). Security Governance Methods and Quantitative Evaluation for Enterprise SMS and Verification Code Systems. In 2026 International Conference on Generative Artificial Intelligence and Information Security (GAIIS) (pp. 455-458). IEEE.
27. Tariq, S., Baruwal Chhetri, M., Nepal, S., & Paris, C. (2025). Alert fatigue in security operations centres: Research challenges and opportunities. ACM Computing Surveys, 57(9), 1-38.
28. Xu, T., Zhu, W., & Zhang, J. (2026). A Study on the Application of Alternative Data in Credit Assessment for the Unbanked Population.
29. Brito, S. D., Azinheira, G. J., Semião, J. F., Sousa, N. M., & Litrán, S. P. (2025). Non-intrusive low-cost IoT-based hardware system for sustainable predictive maintenance of industrial pump systems. Electronics, 14(14), 2913.
30. Xie, Z., Ren, X., Zheng, T., Bai, J., Fan, W., Xu, B., ... & Song, Y. (2026). A Survey on AI Agent Harness. ResearchGate Preprint. DOI, 10.