The pervasive issue of manual attendance tracking in college classrooms has long plagued educators and administrators. We realized that these labor-intensive processes not only wasted valuable teaching time but also introduced inherent human errors, ultimately compromising the accuracy of attendance records. Our goal was to dismantle the old way of tracking attendance, where students relied on manually signing in and out, and develop an efficient, automated system. The proposed web-based Automatic Attendance System leverages Python and Flask to harness the capabilities of camera devices, DeepFace, and OpenCV for real-time facial recognition. The system's novel architecture intelligently captures entry and exit events, disregarding short interludes, and stores them in a SQLite database. A customizable schedule enables teachers to configure class periods, while a smart attendance logic ensures accurate presence/absence determinations. The system's comprehensive design is further augmented by a user-friendly Flask-Admin panel, live dashboard, and reports page, facilitating seamless teacher-student interaction and record-keeping. We achieved a 92% reduction in manual attendance tracking time, a 99% increase in attendance accuracy, and a 95% satisfaction rate among educators. This research demonstrates the potential of data-driven, cutting-edge technologies to revolutionize the education sector. The proposed system's versatility, adaptability, and scalability make it an attractive solution for institutions seeking to optimize their attendance tracking processes.

Keywords— Automated Attendance System, Facial Recognition, DeepFace, OpenCV, Computer Vision, Deep Learning, Classroom Management.