Cat Generation with generative adversarial networks




This cat has been generated with a generative adversarial network (gan). We add a simple tool for replacing the generative model by a better one, using automatic quality assessment.
The original GAN is based on StyleGan2 and we use data from thiscatdoesnotexist.com, thishorsedoesnotexist.com and thisartworkdoesnotexist.com.

Authors: Baptiste Roziere, Fabien Teytaud, Vlad Hosu, Hanhe Lin, Jeremy Rapin, Mariia Zameshina, Olivier Teytaud.


Institutions :

  • Facebook AI Research
  • Univ. Littoral Cote d’Opale
  • Univ. Konstanz, Germany
  • Univ. Grenoble Alpes, CNRS, Inria, Grenoble INP, LIG, Franc

The paper is online here.

@InProceedings{Roziere_2020_ACCV,
    author    = {Roziere, Baptiste and Teytaud, Fabien and Hosu, Vlad and Lin, Hanhe and Rapin, Jeremy and Zameshina, Mariia and Teytaud, Olivier},
    title     = {EvolGAN: Evolutionary Generative Adversarial Networks},
    booktitle = {Proceedings of the Asian Conference on Computer Vision (ACCV)},
    month     = {November},
    year      = {2020}
}

Other examples of image generation with gan

We can generate other fun images, like horses or art pictures: