Format: Virtual Webinar. 45 min. presentation followed by 15 min. Q&A
Please note that 5 sessions for the same lecture will be given at different dates listed below.
Session 2, Thursday, April 16, 2020, 9 am to 10 am IST time (India)
Session 4, Thursday, May 21, 2020, 9 am to 10 am Central Daylight Time (North and South America)
Session 5, Tuesday, May 26, 2020, 4 pm to 5 pm CEST (Central Europe Summer Time)
Five live sessions are completed. Please scroll down to watch the videos from the recordings below. SEG members, view the course for free!
About the course
SEG members, view the course for free! Seismic interpretation involves detecting and extracting structural information, stratigraphic features, and geobodies from seismic images. Although numerous automatic methods have been proposed, seismic interpretation today remains a highly time-consuming task which still requires significant human efforts. The conventional seismic interpretation methods or workflows are not automated or intelligent enough to efficiently or accurately interpret the rapidly increasing seismic data sets, which leaves significantly more data uninterpreted than interpreted.
We improve automatic seismic interpretation by using CNNs (convolutional neural networks) which recently have shown the best performance in detecting and extracting useful image features and objects. One main limitation of applying CNNs in seismic interpretation is the preparation of many training data sets and especially the corresponding geologic labels. Manually labeling geologic features in a seismic image is highly time-consuming and subjective, which often results in incompletely or inaccurately labeled training images. To solve this problem, we propose a workflow to automatically build diverse geologic models with geologically realistic features. Based on these models with known geologic information, we further automatically create numerous synthetic seismic images and the corresponding ground truth of geologic labels to train CNNs for geologic interpretation in field seismic images. Accurate interpretation results in multiple field seismic images show that the proposed workflow simulates realistic and generalized geologic models from which the CNNs effectively learn to recognize real geologic features in field images.
In this lecture, I would like the share you with our research experience on the following topics:
- Automatic preparation of training data sets and labels;
- CNN for fault detection, fault orientation estimation, and fault surface construction;
- CNN for relative geologic time and seismic horizons;
- CNN for seismic geobody tracking;
- CNN-based multitask learning in seismic interpretation;
- CNN for seismic imaging and inversion.
Xinming Wu joins the USTC (University of Science and Technology of China) as a professor in 2019, where he is starting the Computational Interpretation Group (CIG).
Xinming received an engineering degree (2009) in geophysics from Central South University, an M.Sc. (2012) in geophysics from Tongji University, and a Ph.D. (2016) in geophysics from the Colorado School of Mines where he was a member working with Dave Hale at the Center for Wave Phenomena. He interned twice at Transform Software and Services/DrillingInfo during the summer and winter of 2014. From 2016 to 2019, he was a postdoctoral fellow working with Sergey Fomel at Bureau of Economic Geology, The University of Texas at Austin.
He received SEG awards for Best Paper in Geophysics with Dave Hale in 2016 (3D seismic image processing for faults), Best Student Poster Paper with Sean Bader and Sergey Fomel in 2017 (Missing log interpolation and seismic well ties), and an Honorable Mention for Best Paper presented at the 2018 SEG Annual Meeting with Sergey Fomel in 2018 (Least-squares horizons).
He also received the Shanghai excellent master thesis award in 2013 (Generating 3D seismic Wheeler volumes: methods and applications).
Xinming writes a lot of software packages for his research on seismic structural and stratigraphic interpretation, deep learning (e.g., FaultSeg), subsurface modeling, joint seismic and well-log interpretation, and geophysical inversion with geologic constraints.
Start1. Introduction (14:23)
Start2. Training datasets for deep learning (19:43)
Start3. Deep learning for seismic horizon and fault interpretation (8:31)
Start4. Deep learning for seismic geobody interpretation (4:02)
Start5. Deep learning for seismic image processing (1:23)
Start6. Deep learning for seismic imaging and inversion (2:09)
Start7. Conclusion and Acknowledgements (2:19)
Start1. What is deep learning (17:03)
Start2. Introduction (15:48)
Start3. Training datasets for deep learning (13:50)
Start4. Deep learning for seismic horizon and fault interpretation (5:56)
Start5. Deep learning for seismic geobody interpretation (5:44)
Start6. Deep learning for seismic image processing (5:06)
Start7. Deep learning for seismic imaging and inversion (3:45)
Start8. Conclusion and acknowledgements (8:55)