3D simulations of earth textures using generative adversarial networks and 2D examples. Collection of pretrained models, training images and output samples accompagning the DiAGAN repository : https://github.com/randlab/DiAGAN
Repository accompagning this article :
3D Geological Image Synthesis From 2D Examples Using Generative Adversarial Networks
Guillaume Coiffier, Philippe Renard, Sylvain Lefebvre
Frontiers in Water, 28 October 2020
https://www.frontiersin.org/journals/water/articles/10.3389/frwa.2020.560598/full