Zero trust biometric attendance: A secure face recognition framework

Automated attendance systems using face recognition present significant privacy challenges that require urgent attention due to their widespread adoption in academic and corporate environments. This research develops and evaluates a secure attendance system that implements AES-256 encrypted biometric storage in Database, addressing critical vulnerabilities in conventional approaches. The proposed solution combines hybrid encryption (AES+Fernet) with dynamic initialization vector generation and role-based access control to ensure GDPRcompliant data handling. Through rigorous testing, the system achieves 98.2% recognition accuracy with 290ms average processing time while reducing privacy risks by 89% compared to unencrypted systems. The architecture prioritizes three key aspects: (1) computational efficiency for real-time deployment, (2) robust security through multi-layered encryption, and (3) practical implementation simplicity. By comparing various encryption strategies and storage approaches, this study identifies optimal configurations that balance performance with privacy protection. The findings demonstrate that proper cryptographic implementation can maintain high recognition accuracy while eliminating common biometric data vulnerabilities. This research provides valuable insights for both system administrators and security practitioners, establishing a framework for developing privacy-preserving attendance systems. The results highlight the feasibility of implementing military-grade encryption without compromising operational efficiency, offering actionable guidelines for organizations transitioning from traditional attendance methods. Furthermore, the study underscores the importance of continuous security enhancements to address evolving threats in biometric data management.