A basic intro to GANs (Generative Adversarial Networks)

26 Oct 2020 (regularly updated)

In my first blog post on Towards Data Science, featured as an Editor’s Pick and in the Getting Started Column by the editorial board, I try to convey the basics of what I learned about GANs (Generative Adversarial Networks) looking back on a 3-month research internship I did. How do GANs work? Why are they so interesting?

GANs are a deep learning method introduced in 2014. For example, they can create realistic-looking yet entirely computer generated photos of people’s faces. GANs are also behind the controversial deepfake videos. Actually, they can imitate any data distribution (image, text, sound, etc).

You can find my article here (reading it does not require to have a Medium account).