Multi-modal biometric fusion based continuous user authentication for E-proctoring using hybrid LCNN-Salp swarm...

Area of Science:

• Computer Science
• Biometrics
• Cybersecurity

Background:

• Remote proctoring during the COVID-19 pandemic presents significant information retrieval challenges.
• Traditional memory-based authentication is insufficient for secure, real-time applications.
• Multi-modal biometric systems offer superior accuracy and security over unimodal approaches.

Purpose of the Study:

• To propose an efficient continuous biometric user authentication system for online evaluations (CBUA-OE).
• To address challenges in live proctoring and real-time authentication system deployment.
• To enhance data security and customer privacy in live applications.

Main Methods:

• Utilized multi-modal biometric verification (e.g., keystrokes, iris, speech, face).
• Implemented a modified wolf optimization algorithm for feature fusion.
• Employed a hybrid LCNN-Salp swarm optimization classifier for continuous authentication.

Main Results:

• The CBUA-OE system demonstrates robustness and efficiency in continuous user authentication.
• The LCNN-Salp swarm optimization classifier shows consistent performance across various datasets.
• Achieved high accuracy, precision, recall, and F-measure compared to state-of-the-art methods.

Conclusions:

• The proposed CBUA-OE system effectively tackles live proctoring difficulties.
• The system is scalable and usable for real-time tracking applications.
• This research advances secure and reliable continuous biometric authentication.