A Comprehensive Approach to Real-time Attendance Systems: Integrating Face Recognition and Emotion Detection and with Web Technologies

Face recognition and emotion detection have gained widespread application in computer vision where machine learning algorithms play pivotal role in extracting facial features. The features are analyzed for identity recognition and emotional state prediction of human being in real time environment. In this study, we present a real-time attendance system that integrates face recognition with emotion detection, adding a new dimension towards attendance management systems. The system employs a dual-path architecture, utilizing ResNet-50 for facial recognition and Vision Transformer (ViT) for emotion detection, ensuring high performance in both the tasks. We also introduce a custom dataset specifically designed for real-time person recognition and emotion detection, allowing the system to meet the unique demands of practical applications. The proposed architecture achieved accuracy of 0.87 for face recognition and 0.90 for emotion detection, with an AUC-ROC score of 0.96. Web-based integration enhances the accessibility and scalability of the system, facilitating deployment in education, corporate environments, and healthcare. Our experiments demonstrate significant improvements in tracking precision and responsiveness over existing solutions, offering a novel, scalable system that bridges biometric technologies with modern attendance management systems.