As you can most certainly appreciate by now, GANs have wide and varied applications, several of which apply very well to games. One such application is the generation of textures or texture variations. We often want slight variations in textures to give our game worlds a more convincing look. This is and can be done with shaders, but for performance reasons, it is often best to create static assets.
Therefore, in this section, we will build a GAN project that allows us to generate textures or height maps. You could also extend this concept using any of the other cool GANs we briefly touched on earlier. We will be using a default implementation of the Wasserstein GAN by Erik Linder-Norén and converting it for our purposes.
One of the major hurdles you will face when first approaching deep learning problems is shaping data to the form you need. In the original...
