FacekeyID: an adaptive weighted sum score-based fusion framework for continuous user authentication

Abstract User authentication has become a crucial service in the digital world as the threats of impostors continue to emerge. Traditional methods of authentication fail to provide full assurance as they can be bypassed by skilled attackers. Additionally, existing systems suffer from repeated login issues that hamper the user experience. Biometric user authentication has emerged as a potential alternative to other schemes and has proven to provide adequate security. This paper proposes a continuous authentication system using multimodal biometrics based on face and keystroke dynamics. A novel Adaptive Weighted Sum Score Fusion (AWSSF) approach is introduced, which considers environmental factors and the user's profile in addition to the biometrics employed in the decision process. Compared to existing approaches, the proposed system is assessed and determined to be non-intrusive and user-friendly, achieving a 3.02% equal error rate.