Introduction

In the previous chapter, the most traditional neural network architecture was explained and applied to a real-life data problem. In this chapter, we will explore the different concepts of CNNs, which are mainly used to solve computer vision problems (that is, image processing).

Even though all neural network domains are popular nowadays, CNNs are probably the most popular of all neural network architectures. This is mainly because, although they work in many domains, they are particularly good at dealing with images, and advances in technology have allowed the collection and storage of large amounts of images, which makes it possible to tackle a great variety of today's challenges using images as input data.

From image classification to object detection, CNNs are being used to diagnose cancer patients and detect fraud in systems, as well as to construct well-thought-out self-driving vehicles that will revolutionize the future.

This chapter will focus on explaining...