Chapter 16
Building Generative Adversarial Networks
IN THIS CHAPTER
Introducing how neural networks can create credible data
Creating a GAN that can generate handwritten numbers
Presenting image and music applications where GANs shine
Deep learning has turned into a hot technology, and new research produces ever more impressive discoveries all the time. Discoveries always appear at an even faster rate during the Neural Information Processing Systems (NeurIPS) conference (https://neurips.cc/), which serves as the stage for everything related to deep learning. The conference is held every year at a different location around the world (most recently, before this book’s publication, in Montréal, Canada).
The conference always makes new technologies available for people to see, but a few fields have received all the attention. Among the impressive variety of applications and new technologies related to deep learning recently introduced at the conference, here are the ones to pay the most attention to: Natural Language Processing (especially for pretrained embeddings like BERT discussed in Chapter 14); Reinforcement Learning (the topic of the next chapter); and Generative Adversarial ...
Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.
Read now
Unlock full access