Rapid seismic domain transfer: Seismic velocity inversion and modeling using deep generative neural networks

Lukas Mosser, Wouter Kimman, Jesper Søren Dramsch, Steve Purves, Alfredo De la Fuente, Graham Ganssle

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

351 Downloads (Pure)

Abstract

Traditional physics-based approaches to infer sub-surface properties such as full-waveform inversion or reflectivity inversion are time-consuming and computationally expensive. We present a deep-learning technique that eliminates the need for these computationally complex methods by posing the problem as one of domain transfer. Our solution is based on a deep convolutional generative adversarial network and dramatically reduces computation time. Training based on two different types of synthetic data produced a neural network that generates realistic velocity models when applied to a real dataset. The system's ability to generalize means it is robust against the inherent occurrence of velocity errors and artifacts in both training and test datasets.
Original languageEnglish
Title of host publicationProceedings of 80th EAGE Conference and Exhibition 2018
PublisherEuropean Association of Geoscientists and Engineers
Publication date2018
Pages487-491
ISBN (Electronic)978-1-5108-7432-9
DOIs
Publication statusPublished - 2018
Event80th EAGE Conference and Exhibition 2018 - Bella Center Copenhagen, Copenhagen, Denmark
Duration: 11 Jun 201814 Jun 2018

Conference

Conference80th EAGE Conference and Exhibition 2018
LocationBella Center Copenhagen
CountryDenmark
CityCopenhagen
Period11/06/201814/06/2018

Fingerprint Dive into the research topics of 'Rapid seismic domain transfer: Seismic velocity inversion and modeling using deep generative neural networks'. Together they form a unique fingerprint.

Cite this