This paper examines geometric facial liveness indicators for secure biometric verification in digital education. Unlike traditional face recognition focused on identity matching, the approach distinguishes a live face from spoofing attempts such as photos, videos, and screen-based imitations. It highlights landmark dynamics, motion consistency, asymmetry, and depth-related changes as useful liveness cues. The study concludes that geometric liveness analysis can improve security in remote learning and online examinations while remaining efficient and interpretable.
Feruza Ortikova (Fri,) studied this question.