Face recognition is essential for many technologies around the world such as video monitoring, interaction between human machines and security systems. As per modern approaches to machine learning, deep learning related techniques give excellent results in terms of accuracy and processing speed in image recognition. Face recognition recommends a modified architecture of the Convolutional Neural Network (CNN) by collecting two standardization operations on 2 layers. The process of normalization which is normalization by batch introduces acceleration of the network. CNN architecture was used to remove distinctive face features and Softmax classifier was used to identify faces within fully connected layer of Convolutional Neural Network

Keywords : Face Recognition, Deep learning Algorithm, Convolutional Neural Network algorithm, Stacked Auto encoder.