It’s a truth universally acknowledged that a Studio Ghibli fan has a deep, deep desire to transport themselves into any of its films, “Blue’s Clues” Skidoo style. And while most of us don’t possess magical powers, we have the next best thing thanks to science.
A Chinese research team from Wuhan University and Hubei University of Technology has created a machine called the “AnimeGAN: A Novel Lightweight GAN for Photo Animation,” which is technology that turns photographs into anime-style art inspired by the likes of Hayao Miyazaki, Makoto Shinkai and Satoshi Kon. The invention uses machine learning through neural style transfer and generative adversarial networks (GANs).
Aside from being a customizable piece for anime fans, AnimeGAN can also help artists and creators by saving time when illustrating lines, textures, colors and shadows like realistic backgrounds.
Folks don’t need to test this out in a lab. A user recently posted AnimeGAN’s open-source code using machine learning program Tensorflow, so that anyone can fulfill those deep-seated fantasies of living in an anime world.
For those who’d rather become an anime waifu, there’s also an app for that. You’re welcome, weebs.
Photos from ITMedia