Chapter 3: Generative Adversarial Network
Generative Adversarial Network, more commonly known as GANs, are currently the most prominent method in image and video generation. As the inventor of the convolutional neural network, Dr. Yann LeCun, said in 2016, "...it is the most interesting idea in the last 10 years in machine learning." The images generated using GANs are superior, in terms of realism, to other competing technologies and things have advanced tremendously since their invention in 2014 by then graduate student Ian Goodfellow.
In this chapter, we will first learn about the fundamentals of GANs and build a DCGAN to generate Fashion MNIST. We'll learn about the challenges in training GANs. Finally, we will learn how to build a WGAN ...
Get Hands-On Image Generation with TensorFlow now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.
