Face Recognition Model Based On MTCNN And Facenet

Face recognition technology is widely used in various fields, such as time and attendance, payment, access control, etc., providing great convenience to life. This paper proposes a face recognition model based on MTCNN and Facenet, as traditional face recognition systems mostly use manual feature setting, which has disadvantages such as low recognition accuracy and slow speed. The MTCNN model consists of three layers of convolutional neural networks, namely P-Net, R-Net and O-Net, which are used to extract faces from images. The Facenet model is used for face feature vector extraction, and Triplet Loss is used as the loss function to determine the similarity of features through the comparison of Euclidean spatial distances. The final face recognition was completed. The experiments were trained using the LFW dataset and achieved an accuracy of 86.0%.