Survey on Face Recognition using an Improved VGGNET Convolutional Neural Network

Many well-known convolutional neural networks are now based on huge training samples and super computers. It makes lot of issue to overcome this, so this research study develops a deep learning algorithm in VGGNET for Figure classifying and put forward a person’s face to recognition system that called Micro Face. Micro Face employs the CASIA Web Face storage for testing and training samples. This study describes that, when differentiate to the initial algorithm, the upgrade algorithm brings down the parameters, and make that image visible in improving the pooling function, and it would increase the more number of convolution networks for the kernels, which not only decrease the weakness on huge training samples and super computers, it can also carry off 96% identification rate with acceptable identification on performance to achieve and determined value.