Generating unrepresented proportions of geological facies using Generative Adversarial Networks

Alhasan Abdellatif, Ahmed H. Elsheikh, Gavin Graham, Daniel Busby, Philippe Berthet

Research output: Contribution to journalArticlepeer-review

6 Citations (Scopus)
30 Downloads (Pure)


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.
Original languageEnglish
Article number105085
JournalComputers and Geosciences
Early online date12 Mar 2022
Publication statusPublished - May 2022


  • Generative Adversarial Networks (GANs)
  • Multipoint geostatistics
  • Stochastic fields

ASJC Scopus subject areas

  • Information Systems
  • Computers in Earth Sciences


Dive into the research topics of 'Generating unrepresented proportions of geological facies using Generative Adversarial Networks'. Together they form a unique fingerprint.

Cite this