Face Identification based on different Convolutional Neural Network Models for Security System: A Review

Abstract

Computer based authentication system is reliable and mature in technology driven world.  When this authentication is based on biometrics then it is more secure. Face is most prominent identifier in biometric system. Many challenges have come in recognition process when environment is unconstrained. In this paper, different pre trained deep learning models like: VGG16, VGG19, ResNet50, MobileNetV2 are applied to the LFW dataset for Face Recognition. These models are pre trained on ImageNet dataset and for LFW dataset, we achieved the accuracy for fine tune model 73.51%, 74.70%, 87.76%, 90.49% respectively. It has been observed that highestaccuracy has been found by the MobileNetV2which is pre trained on ImageNet dataset.