Generating unrepresented proportions of geological facies using Generative Adversarial Networks
This work addresses a domain-specific challenge in geology by enabling the generation of unrepresented facies proportions, which is incremental as it builds on existing GAN methods with new conditioning routines.
The study tackled the problem of generating geological facies with proportions not present in training data using conditional Generative Adversarial Networks (GANs), achieving good geological consistency and strong correlation with target conditions in numerical experiments on binary and multiple facies images.
In this work, we investigate the capacity of Generative Adversarial Networks (GANs) in interpolating and extrapolating facies proportions in a geological dataset. The new generated realizations with unrepresented (aka. missing) proportions are assumed to belong to the same original data distribution. Specifically, we design a conditional GANs model that can drive the generated facies toward new proportions not found in the training set. The presented study includes an investigation of various training settings and model architectures. In addition, we devised new conditioning routines for an improved generation of the missing samples. The presented numerical experiments on images of binary and multiple facies showed good geological consistency as well as strong correlation with the target conditions.