Generative Deep Learning, 2nd Edition

Chapter 3. Variational Autoencoders

Chapter Goals

In this chapter you will:

• Learn how the architectural design of autoencoders makes them perfectly suited to generative modeling.

• Build and train an autoencoder from scratch using Keras.

• Use autoencoders to generate new images, but understand the limitations of this approach.

• Learn about the architecture of the variational autoencoder and how it solves many of the problems associated with standard autoencoders.

• Build a variational autoencoder from scratch using Keras.

• Use variational autoencoders to generate new images.

• Use variational autoencoders to manipulate generated images using latent space arithmetic.

In 2013, Diederik P. Kingma and Max Welling published a paper that laid the foundations for a type of neural network known as a variational autoencoder (VAE).¹ This is now one of the most fundamental and well-known deep learning architectures for generative modeling and an excellent place to start our journey into generative deep learning.

In this chapter, we shall start by building a standard autoencoder and then see how we can extend this framework to develop a variational autoencoder. Along the way, we will pick apart both types of models, to understand how they work at a granular level. By the end of the chapter you should have a complete understanding of how to build and manipulate autoencoder-based models and, in particular, how to build a variational autoencoder from scratch to generate images based ...