Face Recognition Implementation Using Convolutional Neural Networks on Electronic Guestbook

This research presents a novel approach to streamline visitor registration processes using a convolutional neural network (CNN)-based face recognition system. By capturing and storing facial features in a database, the system eliminates the need for repetitive data entry, significantly reducing manual input time and enhancing data accuracy. The proposed system utilizes two CNN models, each for face detection and face embedding, and Euclidean Distance metric for face identification. The system was evaluated on a primary and a secondary test data. Testing results demonstrate high performance, with an average time reduction of 54.32 seconds and accuracy rates of 100% for primary test data and 87% for secondary test data, respectively. The implementation of this face recognition system offers several benefits. It enhances visitor experiences by eliminating the need for lengthy registration procedures, improves administrative efficiency by reducing manual data entry, and ensures the reliability of visitor data. This innovative solution has the potential to be applied across various organizations, from corporate offices to educational institutions and public venues.