1887

Abstract

Summary

Machine learning approaches are rapidly finding their way into many applications in processing and imaging seismic data. More specifically, various convolutional deep-learning architectures are currently being explored for seismic data processing tasks from denoising to imaging. Here, we present Recurrent Inference Machines (RIMs): a recurrent network architecture designed specifically for inverse problems, where a known forward operator is known and used as a constraint. We describe how both the original RIM and its invertible counterpart (iRIM) are designed to mimic gradient-based optimisation methods, and thus learn to perform data-driven regularisation and implicit model shaping due to their deep learning nature. We show examples of using RIMs to perform seismic data interpolation and image-domain inversion by deblurring, benchmarking them against UNets as a more widely-used deep learning architecture. Our examples show that RIMs outperform UNets particularly in dealing with features not necessarily present in the training data, due to the role of the forward operator as an additional constraint in training.

Loading

Article metrics loading...

/content/papers/10.3997/2214-4609.202211146
2022-06-06
2024-03-28
Loading full text...

Full text loading...

References

  1. [1]KuijpersD., VasconcelosI., andPutzky P. [2021] Reconstructing missing seismic data using Deep Learning. arXiv preprint.
    [Google Scholar]
  2. [2]FletcherR. P., ArcherS., NicholsD., andMao W. [2012] Inversion after depth imaging. SEG Technical Program Expanded Abstracts.
    [Google Scholar]
  3. [3]PutzkyP., andWelling M. [2017] Recurrent inference machines for solving inverse problems. arXiv preprint.
    [Google Scholar]
  4. [4]PutzkyP., andWelling M. [2019] Invert to Learn to Invert.Advances in Neural Information Processing Systems.
    [Google Scholar]
  5. [5]PutzkyP., KarkalousosD., TeuwenJ., MiriakovN., BakkerB., CaanM., andWelling M. [2019] i-RIM applied to the fastMRI challenge. arXiv preprint.
    [Google Scholar]
http://instance.metastore.ingenta.com/content/papers/10.3997/2214-4609.202211146
Loading
/content/papers/10.3997/2214-4609.202211146
Loading

Data & Media loading...

This is a required field
Please enter a valid email address
Approval was a Success
Invalid data
An Error Occurred
Approval was partially successful, following selected items could not be processed due to error