ERL FISH: Nan You (Purdue U.): 3D Carbonate Digital Rock Reconstruction using Progressive Growing GAN

Dr. Nan You, Postdoctoral Scholar at Purdue U., presents “3D Carbonate
Digital Rock Reconstruction using Progressive Growing GAN,” hosted
virtually by MIT Earth Resources Laboratory.

“Digital Rock Physics relies on the availability of high-resolution,
large-size 3D digital rock images. In practice, there is always a
trade-off between the size and resolution of the acquired images.
Moreover, it is time-consuming to acquire high-quality digital rock
images using imaging techniques like X-ray micro-Computed Tomography
(mCT) and Scanning Electron Microscope (SEM). In this paper, we propose
an ML-aided 3D reconstruction method that allows to reduce the sampling
rate along the axial direction during image acquisition.

The key idea is to train a Progressive Growing Generative Adversarial
Network (PG-GAN) to generate high-quality gray-scale cross-section
images and then to reconstruct the 3D digital rock by linearly
interpolating the inverted latent vectors corresponding to the sparsely
scanned cross-section images. We apply our method to an Estaillades
carbonate rock sample. We observe that both the reconstructed image and
the extracted pore network are visually indistinguishable from the
ground truth. Overall, our method achieves nine times speedup of the
imaging process and over 4,500 times compression of the image data for
the Estaillades carbonate rock sample. Moreover, the PG-GAN can enlarge
the digital rock repository and enable efficient imaging editing in its
linear latent space.”

Dr. Nan You is a Postdoc from the Department of Earth, Atmospheric,
and Planetary Sciences at Purdue University. She obtained her B.Eng. in
hydrogeology from Nanjing University in 2016 and her Ph.D. in Geophysics
from National University of Singapore in 2021. She has been working on
ML-based interpretation and integration of different data types (e.g.,
digital rocks, lab data, and well logs) for comprehensive and efficient
rock characterization. She worked with scientists from Halliburton and
ExxonMobil to develop real-time automatic ML tools for well-log
processing and interpretation. Her paper, “3D carbonate digital rock
reconstruction using progressive growing GAN” (You, N., Y. E. Li, and A.
Cheng, 2021), was selected in Eos Editors’ Highlight.

The Earth Resources Laboratory (ERL) Friday Informal Seminar Hour presents guest speakers on topics of interest to our lab including seismology, geophysics, rock physics, machine learning, and the energy industry. Contact fish_seminar_organizers@mit.edu for more information and Zoom password.