Vol. 11, Issue 6, Part D (2025)

Abstract
Face detection is a crucial component of computer vision with applications ranging from biometric authentication and surveillance to social media and human-computer interaction. The primary objective of face detection is to accurately identify and localize human faces within digital images or video streams. Over the years, face detection has evolved from traditional hand-crafted feature-based methods to highly robust and efficient deep learning-based models. This shift has been driven by the need for higher accuracy, better generalization to diverse real-world conditions, and real-time processing capabilities. This review paper presents a comprehensive survey of face detection techniques, with a specific focus on advancements powered by deep learning. The paper begins with an overview of classical methods including Viola-Jones, HOG-SVM, and landmark-based detectors. It then delves into deep learning models such as Convolutional Neural Networks (CNNs), Region-based CNNs (R-CNNs), Single Shot Detectors (SSDs), YOLO (You Only Look Once), and Multi-task Cascaded Convolutional Networks (MTCNN), highlighting their architectures, performance, and deployment efficiency. Additionally, we explore the datasets commonly used for training and evaluation, along with the key performance metrics that benchmark model accuracy and robustness. The paper also outlines various real-world applications, current challenges like occlusion and lighting variation, and discusses future research directions including transformer-based detection and edge AI. By systematically reviewing the evolution and current state of face detection technologies, this paper aims to serve as a valuable resource for researchers, developers, and practitioners interested in the intersection of deep learning and face analytics.