![](\"https:\/\/cdn0.tnwcdn.com\/wp-content\/blogs.dir\/1\/files\/2020\/10\/human2Zombie.jpg\")
The app\u2019s<\/a> creator Jose Brown Kramer built the system by first transfer-learning a StyleGAN2 zombie generator. He then tweaked it to develop a hybrid model comprised of earlier layers from a human generator and later ones from the zombie generator.<\/p>\n [Read: What audience intelligence data tells us about the 2020 US presidential election<\/a>]<\/span><\/em><\/p>\n \u201cI dumped 50,000 pairs of images \u2014 the first from the human StyleGAN2 generator, and the second with the same latent space representation, but passed through the zombie generator,\u201d Kramer explained on Reddit<\/a>. \u201cI then used Pix2PixHD to learn a mapping from the pairs.\u201d<\/span><\/p>\n Kramer says the system stores no record of your image, so your transformation into a zombie should be fairly painless.<\/p>\n The results are pretty impressive. Just check out this terrifying creature:<\/p>\n But if the glut of hackneyed zombie movies has put you off walking corpses, you could instead try an AI-powered transformation into an evil doll<\/a>.<\/p>\n Watch the video below to see another handy old man transform into a demon spawn of Chucky.<\/p>\n Happy Halloween!<\/p>\n Published October 27, 2020 \u2014 18:13 UTC <\/p>\n![](\"https:\/\/cdn0.tnwcdn.com\/wp-content\/blogs.dir\/1\/files\/2020\/10\/human2Zombie-3.jpg\")