Recent advancements in face recognition technology have revolutionized biometric authentication, empowering educational institutions, corporate offices, and other organizations to boost operational efficiency. This paper offers an extensive review of face recognition methodologies, spanning traditional approaches such as Principal Component Analysis (PCA) paired with Local Binary Patterns (LBP) to cutting-edge Convolutional Neural Networks (CNNs) like FaceNet. When combined, these techniques achieve exceptional accuracy, reaching up to 99% across varied environmental conditions. Despite their robustness against challenges such as pose variations, illumination changes, and occlusions, these systems face limitations. Persistent high computational demands, susceptibility to adversarial attacks, dataset biases, and privacy concerns remain critical hurdles. To address these, this study assesses lightweight edge-computing platforms and corporate solutions, while investigating fairness strategies to mitigate system-induced discrimination. Drawing on real-time performance analysis from case studies, this work highlights pathways for developing explainable AI frameworks, energy-efficient architectures, fortified cybersecurity protocols, and multi-biometric systems. Central to this evaluation is the integration of technological innovation with ethical principles, ensuring secure attendance systems meet regulatory demands while cultivating societal trust.