Published Paper Details:

Student attendance tracking is a crucial aspect of academic institutions, ensuring discipline and monitoring student participation. Traditional attendance systems, including manual roll calls and RFID-based methods, are time consuming, prone to human error, and susceptible to fraudulent practices such as proxy attendance. To address these limitations, this research proposes an automated student attendance system utilizing Faster R-CNN (Region-based Convolutional Neural Network) for efficient and accurate face detection, combined with side-angle detection to enhance recognition when students are not directly facing the camera. The system employs a highresolution camera to capture real-time classroom footage. Faster R-CNN is leveraged for fast and precise multi-face detection, ensuring robustness even in large classrooms with multiple students present. However, traditional face recognition models struggle with side-angle or partially occluded faces, leading to misidentification or missed attendance marking. To overcome this challenge, our system integrates a side-angle detection mechanism using deep learning techniques to analyze and classify facial orientations. This mechanism compensates for varying head poses by either applying pose normalization or using angle-aware embedding's, ensuring accurate recognition of students even when they are not facing the camera directly. Once a face is successfully recognized, the system crossreferences it with an existing student database and automatically updates attendance records. The processed data is securely stored in a database, providing real-time access for faculty and administrators. The proposed method significantly improves attendance accuracy, minimizes false negatives, and ensures reliability across different environmental conditions, such as varying lighting and occlusions. Additionally, the system enhances security by preventing unauthorized attendance marking and eliminating the possibility of proxy attendance. This research demonstrates that integrating Faster R-CNN with side-angle detection improves face recognition performance in classroom environments, making it a viable and efficient solution for real-world deployment. The proposed system not only automates attendance tracking but also enhances student monitoring and management in educational institutions, paving the way for intelligent classroom automation.