Published Paper Details:

The non-intrusive nature of face recognition and its broad range of applications in security, surveillance, access control, and identity verification systems have made it one of the most popular biometric technologies. The need for precise and effective real-time face recognition systems has grown dramatically with the quick development of digital imaging equipment, artificial intelligence, and embedded computing platforms. However, differences in lighting, position, facial expression, occlusion, and computing restrictions make it difficult to achieve high identification accuracy while meeting real-time requirements. A thorough analysis of real-time face recognition systems is given in this study, which covers both conventional and modern methods, publicly accessible datasets, performance assessment measures, and deployment issues. In addition to local feature-based approaches like LBP, HOG, SIFT, and their variations, the study methodically examines traditional appearance-based and statistical techniques like PCA, LDA, and ICA. The contributions and limits of key-point descriptors and correlation filter-based techniques in real-time settings are also examined. The study also addresses the shift to deep-learning-based facial recognition models, with a focus on resource-constrained and real-time systems that are light weight. The size, annotation kinds, and appropriateness for real-time assessment of publicly accessible datasets that are often used for benchmarking face recognition systems are summarized. Along with the trade-offs between accuracy, speed, memory usage, and energy efficiency, this study also looks at the role of hardware acceleration employing CPUs, GPUs, and FPGAs. There is also discussion of important issues including scalability, privacy, security, and ethical issues. This study seeks to provide academics and practitioners a comprehensive knowledge of the present status of real-time face recognition and to suggest attractive avenues for future research by compiling recent developments and highlighting outstanding research concerns.