Summary

In this chapter, we described the core concepts of a convolutional neural network in more detail. We provided a summarized history of the CNN and how it originated. We covered the basics of CNNs, ranging from network architecture, layers, loss function, and regularization techniques. We also outlined practical recommendations for each of these concepts and also illustrated how to implement a simple digit classification system using TensorFlow. We also outlined how to use a pre-trained model for a custom application development. Finally, we illustrated popular CNN architectures that often serve as the initial choice of developers for any computer vision task. In next chapter, we will look at how deep learning techniques are applied to the field of natural language processing.