Face Recognition via Convolutional Neural Networks and Siamese Neural Networks

Abstract

Deep convolutional neural networks is playing very important role to solve computer vision task in these decades. In this research has shown implementation state of art face recognition methods and compared them. Advantages and disadvantages of Convolutional and Siamese neural networks is explored for the face recognition task. The novel face recognition system accuracy is checked on the widely used Labelled Faces in the Wild (LFW) dataset. In the study has introduced concepts of triplet loss function. Experiment results can be useful for the developers who is going to create AI applications such as Security system, Self-learning, Visitor analysis system, Face recognition system, Face verification system and many more.