Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{196064,
        author = {Sampath Kumar and Harshith Kataray and Mahin Munawar and CH. Bala Subramanyam},
        title = {A Lightweight Web-Based Framework for Face Recognition–Based Criminal Identity Analysis},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {11},
        pages = {2894-2900},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=196064},
        abstract = {In modern law enforcement and security, the ability to identify individuals in real-time is paramount. However, traditional biometric systems often suffer from high latency and require expensive computational resources, making them difficult to deploy in widespread surveillance networks. This paper introduces a scalable, web-based "Criminal Face Detection System" designed to bridge the gap between high-accuracy forensics and real-time field application. We propose a lightweight client-server architecture that integrates the InsightFace framework, specifically utilizing RetinaFace for robust detection in cluttered backgrounds and ArcFace for generating distinct 512-dimensional facial embeddings. Unlike older statistical methods that struggle with lighting and pose, our approach employs Additive Angular Margin Loss to ensure precise identity separation. The system captures live video via a browser interface and processes data using a FastAPI backend, optimized with the AntelopeV2 model for efficient CPU inference. Preliminary results demonstrate that this modular approach delivers high-fidelity recognition suitable for security checkpoints and criminal identification without the need for heavy GPU infrastructure.},
        keywords = {Face recognition, real-time detection, RetinaFace, ArcFace, InsightFace, facial embeddings, cosine similarity, FastAPI backend, webcam-based recognition, deep learning, computer vision, identity verification, surveillance systems, criminal detection, feature extraction, lightweight model deployment, web application, OpenCV preprocessing.},
        month = {April},
        }