Curtin/CSIRO Geophysics Seminar, 2nd June 2022

Implementation of deep learning algorithms for digital rock analysis

Date: Thursday, 16^nd of June, 2022
Time: 11:00 AM – 12:00 PM
Location: CSIRO/ARRC Auditorium, 26 Dick Perry Avenue, Kensington

Presenter: Dr. Alexandra Roslin (University of Queensland)

Nowadays, deep learning algorithms are widely used for many geosciences applications, including digital rock analysis. For example, deep learning may increase the accuracy of image segmentation or perform rock type classification, enhance the resolution of the micro-CT images and suppress the background noise. Combining deep learning algorithms and machine learning clustering may help predict rock properties based on wellbore data or rock texture depending on reservoir conditions. Deep learning is a growing area of research, and it will find more and more applications in future.

Short biography

Dr. Alexandra Roslin is a postdoctoral research fellow at the University of Queensland and visiting research fellow at Curtin University. She graduated from Novosibirsk State University with a Bachelor of Science (Geology) degree. After working for several years in the oil and gas service industry, she moved to Australia to work in unconventional gas reservoirs. She graduated from the University of Queensland with an MPhil (Earth Sciences) degree and the University of Aberdeen with a PhD (Petroleum Engineering) degree. Alexandra’s current work focuses on implementing deep learning algorithms for digital rock analysis (fracture analysis, mineral segmentation, etc.).