<![CDATA[Face recognition has become an important technology in computer vision and is widely applied in various fields, including identity verification in educational environments. However, variations in lighting, facial pose, and facial expressions remain major challenges in achieving accurate recognition. This study aims to implement and evaluate the performance of a pretrained Convolutional Neural Network (CNN) for student face recognition using a dataset collected directly from ITB Widya Gama Lumajang. The research employed the pretrained CNN model provided by the face_recognition and dlib libraries. The implementation process consisted of image preprocessing, face detection, face encoding, and face matching. System performance was evaluated using accuracy, precision, recall, F1-score, and a confusion matrix. The results demonstrated that the pretrained CNN model achieved an overall recognition accuracy of 96.73% and consistently produced high precision, recall, and F1-score values. Although variations in lighting, facial pose, facial expressions, and image quality affected recognition performance, the system was able to identify most student faces accurately. These findings indicate that the pretrained CNN method is effective and reliable for student face recognition and has the potential to support identity verification applications in educational environments.]]>
Copyrights © 2026
