Deep learning based face detecti̇on and recognition

Face detection and recognition are the most popular research topics that have been working in computer vision for many years. In the literature, there are statistical based, appearance based and template matching based methods for face detection and recognition. However, in most case these methods do not provide the desired success in exposure change, lighting effect and noisy images. In recent years, with the elimination of hardware constraints, the number of layers of artificial neural networks has been increased and deepened models have been developed. These structures are generally known as "deep learning". The machine is also used in almost all of the learning topics in order to solve deep learning on image processing problems that improve the existing situation. Consequently, in this thesis, deep learning which is recently the most popular subject in machine learning has been examined and face detection and recognition with the help of CNN (Convolutional Neural Network) have been realized.