Facelog: Login System with User Authentication Toolkit Utilizing Convolutional Neural Network Algorithm

Abstract

The study introduces FaceLog, a two-layer authentication framework developed to add digital security by incorporating biometric authentication and multi-factor authentication (MFA). The first security layer utilizes a Convolutional Neural Network (CNN)–based facial recognition model with liveness detection to verify user authenticity in real time. Using the Eye Aspect Ratio (EAR) method, the system detects natural eye blinks to distinguish live users from spoofing attempts involving static or digital images. Once facial verification is successful, the system proceeds to second layer of protection, either a One-Time Password (OTP) or a Time-Based One-Time Password (TOTP) for identity confirmation. This structure ensures that even if one authentication factor is compromised, unauthorized access remains effectively prevented. Evaluation results demonstrate high accuracy, precision, recall, and F1-score, supported by excellent ratings in functionality, usability, and compatibility based the criterion of ISO/IEC 25010 software quality model. The findings affirm that combining biometric authentication with multi-factor verification provides a robust, efficient, and user-centered approach to secure modern login systems, addressing the growing challenges of cybersecurity in digital platforms.