ART
BREEDER
It is a collaborative platform using an AI having two models of antagonistic generative networks, typeGAN and BigGAN, from machine learning. Each of the two networks plays a specific role, one will be the "generator" and the other the "discriminator".
The generator proposes a set of images or objects, and the discriminator will judge the accuracy by comparing the rendering with the stated request. This training process puts the two networks in competition so that they can improve in their proposal, this is called retropopagation. This makes it possible to have a more realistic rendering and close to what you want to achieve
​
The site allows you to create from a remix called "Breeding" using images from the database of other users of the site. We can then create new variations and also cross with other images.
​
​