The human face is an important entity that plays a major role in our daily social uses, like conveying a separate identity. The Face recognition system is also able to recognize the person from a distance without gives any disturbance to the person. For face recognition purposes, we need large data sets and complex features. Large data sets are used to uniquely identify the different subjects. For this we need to manipulating different obstacles like illumination, propose this project we propose a deep unified model for Face Recognition based on Faster Region Convolution Neural Network(RDCNN).In our proposed system, we have a several classrooms of a specific institute in which we set up our face recognition system for making smart classrooms. Several images from different smart classrooms are being sent simultaneously for processing to take the attendance. To measure the validity of our application we used a group-based web application attendance system. Then the staff is not attending the class at the correct time then this system is automatically passing the intimations to consent staff. If staff not attending a class at the correct time our proposed system passes the intimations to alternative staff.