In this chapter, we will explore the challenges and the solutions related to the face recognition problem. Therefore, we are going to first present the face recognition problem nature and the similarity function as the general high-level solution. Then, we will introduce Siamese networks, which, together with the similarity function, constitute the fundamental techniques for solving face detection in an efficient manner. From there, we will proceed with two ways that have shown excellent results in training the convolutional neural network for face detection; the triplet loss function and binary classification. Finally, we will see how to use inception network like GoogLeNet and similar transfer learning and the triplet cost function to build the Java face recognition application. Additionally, we will be going through the code details and building a Java application.

The following are the topics that we will be covering in this chapter:

• Problems in face detection
• Differentiating inputs with Siamese networks
• Exploring triplet loss
• Binary classification
• Building face recognition application