Skip to content

StyleGAN

Ashok Bhat edited this page Sep 19, 2020 · 5 revisions

About

  • A GAN for producing an unlimited number of (often convincing) portraits of fake human faces.
  • Introduced by Nvidia researchers in December 2018.
  • Depends on Nvidia's CUDA software, GPUs and on TensorFlow

An example picture (not a real person)

See also

Clone this wiki locally