Perancangan Sistem Deteksi Wajah Real-Time Menggunakan Convolutional Neural Network pada Perangkat Komputer Desktop

Abstract

This research aims to design and implement a real-time face detection system use a Convolutional Neural Network (CNN) on a desktop computer. The research consisted of data preprocesing, collection, model training, and performance evaluation. The dataset contained 200 facial images, which were resized to 224 ×is 224 pixels, normalized, and divided into training and testing sets. The model was developed using TensorFlow, Keras, and OpenCV, and evaluated use confusion matrix based on precision, accuracy, recall, and F1-score. The research results is that the proposed model can detect and recognize faces effectively, as demonstrated by an accuracy value of 97.22%, along with recall, precision, and F1-score values of 97%.  The implementation of Batch Normalization and Dropout improved training stability and enhanced the model's generalization capability. These findings indicate that the CNN -based approach is effective for real-time face detection on desktop computer systems.