Face Recognition Attendance System using Support Vector Machine

The Face Recognition Attendance System takes attendance marking to the next level by automating the process through face recognition. It addresses the issues found in older methods like manual registers and RFID. By harnessing the power of machine learning and computer vision tools like Dlib and SVM, this system crafts unique face embeddings for each person at the moment of capture, enabling real-time identification. It's built with Python, using the Flask framework and a MySQL database, and comes packed with features like new user registration, real-time monitoring, and the option to export data in CSV format. A custom dataset was put together, featuring 50 images for each user, all created using the BlazeFace model right in the browser. This system achieves an accuracy rate of 98%, along with macro precision and recall scores of 0.99 and 0.98, respectively. This represents a major advancement in recognizing individuals already stored in the database. This model is a contactless and scalable design, that ensures better accuracy, security, and efficiency, making it a preferable choice for schools and workplaces, while also reducing manual errors and preventing fake attendance.