Performance and Challenges of Deep Learning-Based Face Recognition Systems: A Systematic Review

This research is motivated by the rapid development of Artificial Intelligence (AI) and the increasing use of face recognition systems in various fields such as security and education. Face recognition, as part of computer vision, enables systems to identify individuals based on unique facial characteristics. This study aims to review the application of Deep Learning methods in face recognition systems, with a focus on commonly used algorithms, model performance, and existing challenges. This study employs a systematic literature review approach by analyzing various relevant scientific publications. Data were collected from reliable academic sources and analyzed qualitatively to identify research trends, methods, and gaps. The results of the study indicate that Convolutional Neural Network (CNN) and its variants, such as FaceNet and VGG-Face, are the most widely used methods and demonstrate high performance in terms of accuracy and reliability. However, several challenges remain, including the need for large-scale datasets, variations in image conditions, and high computational complexity. This study contributes by providing a comprehensive synthesis of Deep Learning applications in face recognition and identifying research gaps and opportunities for future development to improve system effectiveness and efficiency.