An Approach Of Dismissal Of Noise In Face Recognition Using Convolution Neural Network (CNN)

The motivation for this paper is the trend in deep learning and artificial intelligence. In last few years, there has been lot of advancement in face recognition technique. The problem faced in this technique is noise in the data-set. We have tried to overcome this problem by using a code which enables the webcam and collects the data. Deep learning is used in a order to detect face. Algorithm consists of following procedure: Calculating Integral Image, computation Haar like elements, AdaBoost Learning design, Cascade Filter. The following steps were used for coding. Further, training of algorithm and followed by testing. We need to train the algorithm. Further testing is done. Large part of raw data is used for training the algorithm and a small part of the raw data are used for trail the algorithm. It is useful for security purpose, biometric, it can also be used for collecting attendance. It can be further improved by adding Infrared (IR) temperature sensor. This will help us not only to recognize the face but also provide the body temperature.