Seismic deblending by self-supervised deep learning with a blind-trace network

Abstract

PreviousNext No AccessFirst International Meeting for Applied Geoscience & Energy Expanded AbstractsSeismic deblending by self-supervised deep learning with a blind-trace networkAuthors: Shirui WangWenyi HuPengyu YuanXuqing WuQunshan ZhangPrashanth NadukandiGerman Ocampo BoteroJiefu ChenShirui WangUniversity of HoustonSearch for more papers by this author, Wenyi HuAdvanced Geophysical Technology Inc.Search for more papers by this author, Pengyu YuanUniversity of HoustonSearch for more papers by this author, Xuqing WuUniversity of HoustonSearch for more papers by this author, Qunshan ZhangRepsolSearch for more papers by this author, Prashanth NadukandiRepsolSearch for more papers by this author, German Ocampo BoteroRepsolSearch for more papers by this author, and Jiefu ChenUniversity of HoustonSearch for more papers by this authorhttps://doi.org/10.1190/segam2021-3583662.1 SectionsAboutPDF/ePub ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InRedditEmail AbstractSimultaneous-source acquisition endows seismic imaging with dense and efficient surveying designs. This cost-efficient strategy leads to inevitable interference when multiple source responses are recorded within a short time interval. To enable traditional seismic processing workflows, we need an accurate deblending process to separate the blended signals. Deep learning has been widely used in various seismic processing tasks. However, the disparity of seismic datasets is a challenge to a deep neural network’s adaptiveness (also known as generalization). The scarcity of labeled data in the deblending task also intensifies the overfitting problem. We propose a self-supervised learning method for seismic data deblending and a flexible deblending algorithm for speed-accuracy tradeoff. Using a novel blind-trace network and blending loss function, self-supervised training is deployed directly on the target data without the need for a labeled training dataset. Numerical experiments on several datasets demonstrate the effectiveness and robustness of the proposed method.Keywords: artificial intelligence, deblending, machine learning, neural networks, simultaneous sourcePermalink: https://doi.org/10.1190/segam2021-3583662.1FiguresReferencesRelatedDetails First International Meeting for Applied Geoscience & Energy Expanded AbstractsISSN (print):1052-3812 ISSN (online):1949-4645Copyright: 2021 Pages: 3561 publication data© 2021 Published in electronic format with permission by the Society of Exploration GeophysicistsPublisher:Society of Exploration Geophysicists HistoryPublished: 01 Sep 2021 CITATION INFORMATION Shirui Wang, Wenyi Hu, Pengyu Yuan, Xuqing Wu, Qunshan Zhang, Prashanth Nadukandi, German Ocampo Botero, and Jiefu Chen, (2021), "Seismic deblending by self-supervised deep learning with a blind-trace network," SEG Technical Program Expanded Abstracts : 3194-3198. https://doi.org/10.1190/segam2021-3583662.1 Plain-Language Summary Keywordsartificial intelligencedeblendingmachine learningneural networkssimultaneous sourcePDF DownloadLoading ...