Quote:
Nvidia researchers Tero Karras, Samuli Laine, and Timo Aila posted details of the method to produce completely imaginary fake faces with stunning, almost eerie, realism (here’s the paper).
GANs employ two "dueling" neural networks to train a model to learn the nature of a dataset well enough to generate convincing fakes. When you apply GANs to images, this provides a way to generate often highly realistic still fakes you could use for extremely hard to detect social engineering attacks, especially combined with deep fake videos.