Real-time Face Recognition Using an Enhanced VGGNET Convolutional Neural Network

Biometrics serves as the foundation for security systems and plays a pivotal role in implementing robust security measures. Among various biometric modalities such as iris, face, and fingerprint recognition, facial recognition stands out as the widely recognized and dependable option for identification and verification in security systems. Using VGGNET for classified incorporating face recognition as a biometric method offers immense potential, delivering high recognition rates while maintaining user-friendly operation. In this study, a real-time system was developed to recognize faces in a video stream obtained from a surveillance camera, specifically focusing on the face recognition approach in the presence of partial occlusion. The implemented system incorporated real-time face detection capabilities to enhance its performance. This approach not only mitigates the limitations associated with large training samples and supercomputers but also achieves an impressive 96% identification rate while maintaining acceptable performance levels and determined values.