Generative Adversarial Networks (GANs)
A Short Introduction to GANs in Generative Deep Learning
The battle between two adversaries
Who doesn't like to generate natural-looking and realistic things that have never existed before?
That’s where GANs (Generative Adversarial Networks) come into play. They are a type of neural network architecture that was specially designed to generate natural-looking and more realistic things such as images, videos, sounds, texts, etc.
With this article, you take one step ahead to learn another neural network architecture. Previously, in our Neural Networks and Deep Learning Course, we’ve discussed the following neural network architectures and their Keras implementations.
Generative deep learning models
There are two main types of generative deep learning models: AEs and GANs. GANs excel in generating realistic things which are almost identical to the original ones. Therefore, introducing GANs is considered one of the most important breakthroughs made in the field of deep learning.
