Volume 2 - Issue 2, March - April 2026

With the increasing demand for secure and transparent electoral processes, biometric-based voter authentication has gained significant importance. Traditional voter verification methods relying on manual identity checks are vulnerable to impersonation, duplicate voting, and lack of reliable identity verification. This paper proposes a Secured Biometric Voter Verification System that integrates deep learning–based facial recognition with fingerprint authentication as a secondary verification mechanism. The facial recognition module performs face detection and alignment using pre-trained models, followed by deep convolutional neural network-based feature extraction to generate discriminative 128-dimensional facial embeddings for identity matching. In cases of facial mismatch or poor image quality, fingerprint verification is triggered to maintain authentication reliability and system continuity. The proposed multimodal framework enhances robustness by combining two biometric traits at the decision level, thereby strengthening authentication security compared to unimodal systems. Biometric templates are securely stored in a protected PostgreSQL database to preserve voter privacy and data integrity. The system is implemented as a scalable web-based prototype suitable for real-time deployment in electronic voting environments. Functional validation under controlled conditions demonstrates reliable authentication performance and effective prevention of duplicate voting, highlighting the practical feasibility of multimodal biometric verification in modern digital election infrastructures.

Karthika.S, Karunyashri.S, Suganthi.A, Mahalakshmi.S, Abinaya.M,"Deep Learning-Based Biometric Voter Verification System with Facial Recognition and Fingerprint Fallback Authentication" International Journal of Advanced Multidisciplinary Research and Educational Development, V2(2): Page(1442-1447) Mar-Apr 2026. ISSN: 3107-6513. www.ijamred.com. Published by Scientific and Academic Research Publishing.