Stochastic seismic waveform inversion using generative adversarial networks as a geological prior
From MaRDI portal
Publication:2284094
Abstract: We present an application of deep generative models in the context of partial-differential equation (PDE) constrained inverse problems. We combine a generative adversarial network (GAN) representing an a priori model that creates subsurface geological structures and their petrophysical properties, with the numerical solution of the PDE governing the propagation of acoustic waves within the earth's interior. We perform Bayesian inversion using an approximate Metropolis-adjusted Langevin algorithm (MALA) to sample from the posterior given seismic observations. Gradients with respect to the model parameters governing the forward problem are obtained by solving the adjoint of the acoustic wave equation. Gradients of the mismatch with respect to the latent variables are obtained by leveraging the differentiable nature of the deep neural network used to represent the generative model. We show that approximate MALA sampling allows efficient Bayesian inversion of model parameters obtained from a prior represented by a deep generative model, obtaining a diverse set of realizations that reflect the observed seismic response.
Recommendations
- Towards blending physics-based numerical simulations and seismic databases using generative adversarial network
- Application of Bayesian generative adversarial networks to geological facies modeling
- Parametric generation of conditional geological realizations using generative neural networks
- Fast geostatistical seismic inversion coupling machine learning and Fourier decomposition
- Geological facies modeling based on progressive growing of generative adversarial networks (GANs)
- Seismic waveform inversion based on a Bayesian-sparsity constraint regularization method
- An encoder-decoder deep surrogate for reverse time migration in seismic imaging under uncertainty
Cites work
- Bayesian Gaussian mixture linear inversion for geophysical inverse problems
- Definitions and examples of inverse and ill-posed problems
- Devito
- Exponential convergence of Langevin distributions and their discrete approximations
- Inverse Problem Theory and Methods for Model Parameter Estimation
- Langevin diffusions and the Metropolis-adjusted Langevin algorithm
- Multilayer feedforward networks are universal approximators
- Optimal Scaling of Discrete Approximations to Langevin Diffusions
- Parametric generation of conditional geological realizations using generative neural networks
- Resolution analysis of general inverse problems through inverse Monte Carlo sampling
- Stochastic seismic waveform inversion using generative adversarial networks as a geological prior
Cited in
(34)- Parametric generation of conditional geological realizations using generative neural networks
- Fast geostatistical seismic inversion coupling machine learning and Fourier decomposition
- Towards blending physics-based numerical simulations and seismic databases using generative adversarial network
- Geological facies modeling based on progressive growing of generative adversarial networks (GANs)
- Feature extraction of music signal based on adaptive wave equation inversion
- Reconstruction, optimization, and design of heterogeneous materials and media: basic principles, computational algorithms, and applications
- GANSim: conditional facies simulation using an improved progressive growing of generative adversarial networks (GANs)
- Solution of physics-based Bayesian inverse problems with deep generative priors
- 3D reconstruction of porous media using a batch normalized variational auto-encoder
- Conditioning generative adversarial networks on nonlinear data for subsurface flow model calibration and uncertainty quantification
- Seismic inversion with deep learning. A proposal for litho-type classification
- Bayesian multiscale deep generative model for the solution of high-dimensional inverse problems
- Data-space inversion using a recurrent autoencoder for time-series parameterization
- Recent developments combining ensemble smoother and deep generative networks for facies history matching
- U-net generative adversarial network for subsurface facies modeling
- Seismic stratum segmentation using an encoder-decoder convolutional neural network
- Linear waveform tomography inversion using machine learning algorithms
- Stochastic seismic waveform inversion using generative adversarial networks as a geological prior
- Solving Bayesian inverse problems from the perspective of deep generative networks
- A deep-learning-based geological parameterization for history matching complex models
- Application of Bayesian generative adversarial networks to geological facies modeling
- Enforcing imprecise constraints on generative adversarial networks for emulating physical systems
- A Statistical Framework for Generating Microstructures of Two-Phase Random Materials: Application to Fatigue Analysis
- Sojourn functionals for spatiotemporal Gaussian random fields with long memory
- Data assimilation predictive GAN (DA-PredGAN) applied to a spatio-temporal compartmental model in epidemiology
- Regularising inverse problems with generative machine learning models
- Identification of the flux function of nonlinear conservation laws with variable parameters
- Non-Newtonian flow on homogeneous-heterogeneous pore-scale reactive transport: a computational analysis
- Production well placement and history matching by hyperparametric optimization and machine learning
- Deep learning in computational mechanics: a review
- Neural-network-based regularization methods for inverse problems in imaging
- Integrating physics of the problem into data-driven methods to enhance elastic full-waveform inversion with uncertainty quantification
- Latent space inference via paired autoencoders
- Deep learning seismic substructure detection using the frozen Gaussian approximation
This page was built for publication: Stochastic seismic waveform inversion using generative adversarial networks as a geological prior
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2284094)
