Automated biometric attendance monitoring has emerged as a critical research domain within educational technology, driven by the persistent limitations of conventional attendance methods such as manual roll calls, sign-in sheets, and card-based systems. Recent advances in deep learning and computer vision have substantially transformed this field, enabling the development of robust, contactless, and scalable face recognition systems. This review critically examines the evolution of automated biometric attendance monitoring systems across 34 primary studies, tracing the progression from classical handcrafted feature extraction methods, including Histogram of Oriented Gradients (HOG), Local Binary Patterns (LBP), and Eigenface-based approaches, to contemporary deep convolutional neural network architectures and deep metric learning frameworks such as DeepFace, FaceNet, VGGFace, OpenFace, and ResNet-based embedding models.