Fundamentals of Deep Learning

Chapter 5. Convolutional Neural Networks

Neurons in Human Vision

The human sense of vision is unbelievably advanced. Within fractions of seconds, we can identify objects within our field of view, without thought or hesitation. Not only can we name objects we are looking at, we can also perceive their depth, perfectly distinguish their contours, and separate the objects from their backgrounds. Somehow our eyes take in raw voxels of color data, but our brain transforms that information into more meaningful primitives—lines, curves, and shapes—that might indicate, for example, that we’re looking at a house cat.¹

Foundational to the human sense of vision is the neuron. Specialized neurons are responsible for capturing light information in the human eye.² This light information is then preprocessed, transported to the visual cortex of the brain, and then finally analyzed to completion. Neurons are single-handedly responsible for all of these functions. As a result, intuitively, it would make a lot of sense to extend our neural network models to build better computer vision systems. In this chapter, we will use our understanding of human vision to build effective deep learning models for image problems. But before we jump in, let’s take a look at more traditional approaches to image analysis and why they fall short.

The Shortcomings of Feature Selection

Let’s   begin by considering a simple computer vision problem. I give you a randomly selected image, such as the one in Figure 5-1 ...