Neural networks for geophysicists and their application to seismic data interpretation

Peters, Bas; Haber, Eldad; Granek, Justin

doi:10.1190/tle38070534.1

Cited by 34 publications

(8 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The matrix Q selects from the prediction, the pixel indices where we have labels. This is analogous to restricting physical fields to sensor locations [17,19]. Compared to earlier regularized PDE-constrained optimization formulations [8], we propose to apply αR(y n ) to the network output (and not on the parameters K as in weight-decay, or parameter smoothness in time [8]), such that we can explicitly control the properties of the predicted probability maps.…”

Section: Output-regularized Neural Network Training As Pde-constraine...mentioning

confidence: 99%

Deep connections between learning from limited labels & physical parameter estimation -- inspiration for regularization

Peters

2020

Preprint

Self Cite

View full text Add to dashboard Cite

Recently established equivalences between differential equations and the structure of neural networks enabled some interpretation of training of a neural network as partial-differential-equation (PDE) constrained optimization. We add to the previously established connections, explicit regularization that is particularly beneficial in the case of single large-scale examples with partial annotation. We show that explicit regularization of model parameters in PDE constrained optimization translates to regularization of the network output. Examination of the structure of the corresponding Lagrangian and backpropagation algorithm do not reveal additional computational challenges. A hyperspectral imaging example shows that minimum prior information together with cross-validation for optimal regularization parameters boosts the segmentation accuracy.

show abstract

Section: Output-regularized Neural Network Training As Pde-constraine...mentioning

confidence: 99%

Deep connections between learning from limited labels & physical parameter estimation -- inspiration for regularization

Peters

2020

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…In recent years, machine learning has thus been applied to numerous problems in seismic interpretation, such as: (1) salt detection (e.g., Guillen et al, 2015;Waldeland et al, 2018) , (2) fault detection (e.g., Zhang et al, 2014;Araya-Polo et al, 2017;Wu and Fomel, 2018) , (3) horizon mapping (e.g., Peters et al, 2019;Tschannen et al, 2020) , and (4) seismic facies classification (e.g., Qian et al, 2018;Wrona et al, 2018;Liu et al, 2019) . Most of these studies are built on recent advances in machine learning of multi-layered neural networks (i.e.…”

Section: Introductionmentioning

confidence: 99%

3-D seismic interpretation with deep learning: a brief introduction

Wrona¹,

Pan²,

Bell³

et al. 2022

Preprint

View full text Add to dashboard Cite

Understanding the internal structure of our planet is a fundamental goal of the Earth Sciences. As direct observations are restricted to surface outcrops and borehole cores, we rely on geophysical data to study the Earth's interior. Especially, seismic reflection data showing acoustic images of the subsurface, provide us with critical insights into sedimentary, tectonic and magmatic systems. The interpretation of these large, 2-D grids or 3-D seismic volumes is however time-consuming even for a well-trained person or team of people. Here we demonstrate how to automate and accelerate the analysis of these increasingly large seismic datasets with machine learning. We are able to perform typical seismic interpretation tasks, such as the mapping of (1) tectonic faults, (2) salt bodies and (3) sedimentary horizons at high accuracy using deep convolutional neural networks. We share our workflows and scripts, encouraging users to apply our methods to similar problems. Our methodology is generic and flexible allowing an easy

show abstract

“…[30] proposed generative adversarial networks (GANs) as an alternative to the seismic trace interpolation, which can facilitate automated seismic interpretation workflows. [31] discussed the connection between the deep neural networks and other geophysical inverse problems, and showed their utility in lithology interpolation and horizon tracking. The basic idea of these works is taking a deep neural network to learn a required mapping relation by training data sets.…”

Section: Introductionmentioning

confidence: 99%

A Sequential Iterative Deep Learning Seismic Blind High-Resolution Inversion

Chen

Gao

et al. 2021

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

Seismic blind high-resolution inversion (BHRI) aims at retrieving the high-resolution data to characterize the stratigraphic structures in the case of an unknown seismic wavelet. However, the unknown wavelet and ill-posedness pose a great challenge to the high-resolution inversion. Regularization-based BHRI is an effective approach. However, it is sensitive to the sets of initial values, regularization terms, and regularization parameters, and suffers from computational burden problems. To address these issues, we propose a sequential iterative deep learning method shortened to SIDLM to implement a BHRI in a fast computational speed, which incorporates three learned components to sequentially invert initial high-resolution data, seismic wavelet, and final high-resolution data in an end-to-end fashion. Specifically, to mitigate the influence of initial values, a data-driven network U-Net is adopted to learn an initial highresolution data. Further, the architecture makes use of prior information encoded in the forward operator to build a new general Alternating Direction Method of Multipliers (ADMM)like iterative deep neural network which is instead of the traditional alternating iterative inversion. The proposed ADMMlike network utilizes the convolutional neural networks to learn the proximal operators to solve each sub-problems in alternating iterative inversion. Therefore, all parameters of BHRI, such as the regularization parameters and transform operator, can be implicitly learned from the training data sets in an end-toend fashion, not limited to the form of the penalty function. Finally, the synthetic and field data examples are conducted to demonstrate the effectiveness of the proposed SIDLM.

show abstract

Neural networks for geophysicists and their application to seismic data interpretation

Cited by 34 publications

References 25 publications

Deep connections between learning from limited labels & physical parameter estimation -- inspiration for regularization

Deep connections between learning from limited labels & physical parameter estimation -- inspiration for regularization

3-D seismic interpretation with deep learning: a brief introduction

A Sequential Iterative Deep Learning Seismic Blind High-Resolution Inversion

Contact Info

Product

Resources

About