Introduction

In today's world, the need to maintain the security of information is becoming increasingly important, as well as increasingly difficult. There are various methods by which this security can be enforced (passwords, fingerprint IDs, PIN numbers, and so on). However, when it comes to ease of use, accuracy, and low intrusiveness, face recognition algorithms have been doing very well. With the availability of high-speed computing and the evolution of deep convolutional networks, it has been made possible to further increase the robustness of these algorithms. They have gotten so advanced that they are now being used as the primary security feature in many electronic devices (for example, iPhoneX) and even banking applications. The goal of this chapter is to develop a robust, pose-invariant face recognition algorithm for use in security systems. For the purposes of this chapter, we will be using the openly available MIT-CBCL dataset of face images of 10 different subjects.