<![CDATA[This research implements a face recognition-based attendance system using Convolutional Neural Networks method and FaceNet model. This topic was chosen because face recognition is an effective identification method for attendance applications, but often faces challenges of low illumination and varying object distances, especially on devices with mid-to-low specifications. This system uses Convolutional Neural Networks for facial feature extraction, FaceNet to improve face representation accuracy, and Local Binary Patterns Histogram to analyze facial texture to improve recognition performance. The steps taken include collecting face datasets, applying Convolutional Neural Networks and FaceNet models, and evaluating the system under low lighting conditions and various object distances. The test results showed 100% accuracy with three face images even in low lighting conditions. The system still performs well despite variations in light intensity and object distance. The main contribution of this research is the development of an efficient face recognition system based on Convolutional Neural Networks and FaceNet that can be applied to devices with limited specifications for attendance applications, with a focus on stability in poor lighting and testing in real environments.]]>
Copyrights © 2025
