CS SEMINAR:Real-world applications of Deep Learning algorithms...
Title: Real-world applications of Deep Learning algorithms to solve Computer Vision tasks in Geoscience
Date/Time: August 17,2023 1:40PM
Place: FENS G035 (Physical Only)
Abstract: Artificial intelligence (AI) has become an integral part of many industries, and geoscience is no exception. While Australia’s mining industry has delivered a 14.6 percent share of the Australian economy, making it the largest economic contributor, the process for geologists to analyze geoscientific data to generate meaningful information in the exploration and mining industry remained labor, time, and money-consuming. As deep learning improves machine learning and computer vision, the application of ever-evolving algorithms in object detection, segmentation, and classification allows us to automate these works, otherwise manually done by geologists. In this talk, we will explore how these algorithms can be applied as semantic and instance segmentation of valuable minerals; detection, and analysis of fractures; and even detection of a potential new meteorite impact site on the continent of Australia.
Bio:After studying computer science at Izmir Institute of Technology (Turkey), Gokcen received her MS at Bilkent University (Turkey) and her Ph.D. at ETH Zurich (Switzerland) in collaboration with Disney Research Zurich. During her Ph.D. studies, she worked on projects involving, on the one hand, 3D character animation in augmented- and mixed-reality, on the other hand, combining graphics with vision. In 2019, Gokcen joined Datarock in Melbourne, Australia, as a computer vision specialist. Since then, she has been working there, applying machine learning and deep learning workflows to various geoscientific problems in the domain of geology and mining.