We investigate conditional adversarial networks as a general-purpose solution to image-to-image translation problems. These networks not only learn the mapping from input image to output image, but also learn a loss function to train this mapping.
this looks really important to 3d and 2d pipelines
There has to be a trick to this (ie. using a library of images as a reference), there’s no way the algorithm can just conjure up ultra-realistic images from a bunch of cutouts and line drawings.
As it looks more like a foray into new algorithms for AI, I don’t see at the moment just how this will help 3D artists.
Blenderalready has this.
Video title says “april’s fools joke”…come on…
