Abstract
Deep learning has recently drawn massive attention for the task of geophysical inversion because of its strong non-linear mapping ability, especially for seismic inversion. Many works have been proposed. However, to simplify the problem, most of researchers only consider the simple observation setup and regard the task as the directive mapping between seismic data and velocity model, which limits the application in the real scenario. To solve this problem, we adopt the recent proposed SeisInvNet, which reconstructs the velocity model by treating each seismic trace as the essential element. Furthermore, we further extend its encoder part to make it accept randomly sampled traces as input so that it can be suitable for the inversion of seismic data under various observation setups. By comparing with the inversion results, we find that the modified SeisInvNet not only can adapt to various observation setups but also performs slightly better than the original method.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: IOP Conference Series: Earth and Environmental Science
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
