In this chapter, we created a face recognition system based on a Siamese neural network. The face recognition system uses a webcam to stream frames from a live video to a pre-trained Siamese neural network, and using a true image of the user, the system is able to authenticate the user in front of the webcam.

We first dissected the face recognition problem into smaller subproblems, and we saw how a face recognition system first performs a face detection step to isolate the face from the rest of the image, before the actual face recognition step. We saw how face detection is commonly done by the Viola-Jones algorithm, which uses Haar features to detect faces in real time. Face detection using Haar filters is implemented in Python via the OpenCV library, which allows us to perform face detection in just a few lines of code.

We then focused on face recognition, and we discussed...