Real-Time Violence Detection In Surveillance Systems Using Lightweight Hybrid Deep Learning Architectures
Real-Time Violence Detection, MobileNetV2, Bidirectional LSTM (Bi-LSTM), Real Life Violence Situations Dataset, Deep Learning
The exponential growth of surveillance camera networks has created a critical need for automated anomaly detection systems. Traditional manual monitoring is labor-intensive and prone to human error. While recent deep learning models, such as KianNet (ResNet50 combined with ConvLSTM), have achieved high accuracy in detecting violence, they often suffer from high computational costs, making them unsuitable for real-time deployment on edge devices. This paper proposes a lightweight, real-time violence detection system utilizing a hybrid architecture of MobileNetV2 and Bidirectional Long Short Term Memory (Bi-LSTM). By leveraging transfer learning and optimizing input dimensionality to 64 × 64 pixels, our model significantly reduces inference latency while maintaining robust classification performance. Experimental results on the Real Life Violence Situations dataset demonstrate the model’s efficiency, achieving a favorable trade-off between accuracy and speed, thereby bridging the gap between high-performance server-side models and practical real-time surveillance applications. The proposed model demonstrates strong convergence behavior, achieving near-perfect training accuracy and consistently high validation accuracy, indicating effective learning and good generalization performance without significant overfitting.
Registration ID: IJVRA_701906 Published ID: IJVRA2603485
"Real-Time Violence Detection In Surveillance Systems Using Lightweight Hybrid Deep Learning Architectures", IJVRA - International Journal of Versatile Research and Analysis (www.IJVRA.org), ISSN:2984-8903, Vol.4, Issue 3, page no.702-709, March-2026, Available :https://ijpub.org/ijvra/papers/IJVRA2603485.pdf
Paper Reg. ID: IJVRA_701906
Published Paper Id: IJVRA2603485
Research Area: Other area not in list
Country: Markapur, Andhra Pradesh, India
ISSN: 2984-8903 | IMPACT FACTOR: 9.12 Calculated By Google Scholar | ESTD YEAR: 2023
An International UGC CARE JOURNAL PUBLICATION Low Cost (₹599), Scholarly Open Access, Peer-Reviewed, Refereed Journal Impact Factor 9.12 Calculate by Google Scholar and Semantic Scholar | AI-Powered Research Tool, Multidisciplinary, Monthly, Multilanguage, Crossref DOI Member Journal Indexing in All Major Database & Metadata, Citation Generator
Publisher: IJVRA (IJ Publication) Janvi Wave
© 2026 - Authors hold the copyright of this article. This work is licensed under a Creative Commons Attribution 4.0 International License and The Open Definition.
This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (CC BY 4.0). 🛡️ Disclaimer:
The content, data, and findings in this article are based on the authors’ research and have been peer-reviewed for academic purposes only. Readers are advised to verify all information before practical or commercial use. The journal and its editorial board are not liable for any errors, losses, or consequences arising from its use.
: Click Here to Get
About Hard Copy and Transparent Peer Review Report
