This is an academic project, developed as part of my thesis, "Using Generative Adversarial Networks for Content Generation in Games",
in partial fulfillment of the requirements for the degree of MSc in Computer Games Technology.
For the scope of this project I've trained a GAN model using Minecraft data,
in order to generate Minecraft worlds using semantically labeled images.
The model is based on SPADE, a GAN model developed by Nvidia. An example of
what can be achieved with semantic labeling can be seen at
GauGAN.
The training data was kindly provided by Microsoft, which holds intellectual property over it.
For this sensible reason, this project's repository is not available, but the means to achieve
the results I've obtained have been described thoroughly in the dissertation.
The objective of this project was to evaluate the possibility of game world generation using a GAN.