Seismic Data Reconstruction Using Deep Bidirectional Long Short-Term Memory With Skip Connections

Yoon, Daeung; Yeeh, Zeu; Byun, Joongmoo

doi:10.1109/lgrs.2020.2993847

Cited by 56 publications

(11 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…RNNs can use the previous outputs, which are regarded as current inputs, as well as the internal states (memories) to process variable length sequences of inputs. RNN has also been applied to deal with seismic data reconstruction [52], normal moveout velocity estimation [53], and impedance inversion [54]. Thanks to the AD [55] of DSL and the fact that the feedforward computation graph of an RNN seems like the iterative process of time-domain wave propagation, FWI can be well implemented by a suitable RNN.…”

Section: Unsupervised Learning Approaches For Fwimentioning

confidence: 99%

Revisit Geophysical Imaging in A New View of Physics-informed Generative Adversarial Learning

Yang,

2021

Preprint

View full text Add to dashboard Cite

Seismic full waveform inversion (FWI) is a powerful geophysical imaging technique that produces high-resolution subsurface models by iteratively minimizing the misfit between the simulated and observed seismograms. Unfortunately, conventional FWI with least-squares function suffers from many drawbacks such as the local-minima problem and computation of explicit gradient. It is particularly challenging with the contaminated measurements or poor starting models. Recent works relying on partial differential equations and neural networks show promising performance for two-dimensional FWI. Inspired by the competitive learning of generative adversarial networks, we proposed an unsupervised learning paradigm that integrates wave equation with a discriminate network to accurately estimate the physically consistent models in a distribution sense. Our framework needs no labelled training data nor pretraining of the network, is flexible to achieve multi-parameters inversion with minimal user interaction. The proposed method faithfully recovers the well-known synthetic models that outperforms the classical algorithms. Furthermore, our work paves the way to sidestep the local-minima issue via reducing the sensitivity to initial models and noise. Keywords Seismic full waveform inversionRecently, deep-learning (DL) techniques [7], in particular, convolutional neural network (CNN) [8], have achieved the state-of-the-art performance in a variety of applications. They surpass the conventional approaches in many research fields of inverse problem, for instance, image reconstruction [9, 10], x-ray computed tomography [11], and optical diffraction tomography [12,13]. To some extent, this development results from the wide availability of domain-specific languages (DSL) for DL, such as Tensorflow [14] or PyTorch [15], used in academia and industry.

show abstract

Section: Unsupervised Learning Approaches For Fwimentioning

confidence: 99%

Revisit Geophysical Imaging in A New View of Physics-informed Generative Adversarial Learning

Yang,

2021

Preprint

View full text Add to dashboard Cite

show abstract

“…Applications of RNN and LSTM to solve geophysical problems are recent and date back to the very last years. These examples refer to earthquake classification (Kuyuk & Susumu, 2018), detection of earthquake precursors (Cai et al ., 2019), earthquake magnitude prediction (Gonzales et al ., 2019), facies classification from post‐stack seismic data (Grana et al., 2020), seismic velocity analysis (Fabien‐Ouellet & Sarkar, 2020), well log generation (Zhang et al ., 2018), well production prediction (Jie et al ., 2020), seismic data interpolation (Yoon et al ., 2020) and porosity estimation from well log data (Chen et al ., 2020).…”

Section: Introductionmentioning

confidence: 99%

Elastic properties and litho‐fluid facies estimation from pre‐stack seismic data through bidirectional long short‐term memory networks

Aleardi

2022

Geophysical Prospecting

View full text Add to dashboard Cite

One of the main goals of pre‐stack seismic inversion is the estimation of elastic properties (i.e. P‐, S‐ wave velocities and density) and litho‐fluid classes in the investigated area. To this end, many inversion strategies have been proposed, but the most popular is based on a two‐step inversion approach: First, elastic properties are inferred from pre‐stack data, and then a classification algorithm is used to convert the outcomes of the first stage into litho‐fluid facies. In this work, we propose an alternative approach based on recurrent neural networks. We train two bidirectional long short‐term memory networks to predict the inverse mappings from pre‐stack seismic data to elastic properties, and litho‐fluid classes. In the elastic inversion, we also use a Monte Carlo simulation approach to properly propagate onto the model space both the uncertainties related to noise contamination in the data and to the modelling error introduced by the network approximation. One crucial aspect of any machine‐learning inversion strategy is the definition of an appropriate training set. In this case, the models forming the training and validation examples are drawn according to a previously defined elastic and facies prior models derived from actual well log recordings. In particular, we assume a Gaussian‐mixture elastic prior, and we also take into account the uncertainties affecting the estimation of the transition probabilities of facies. We invert each seismic gather independently, and in this context, the generation of the training set, and the learning process can be accomplished with a very limited computational effort on a common notebook. Once trained, the networks estimate the elastic properties, the litho‐fluid facies and the related uncertainties from the pre‐stack data in near real‐time. Synthetic and field data inversions are used to validate the proposed method. The network predictions are also benchmarked against the outcomes of a more standard two‐step approach that combines a linear elastic inversion and a subsequent point‐wise Bayesian classification. Our results demonstrate that the implemented algorithm guarantees more accurate elastic property estimations and facies predictions than the standard inversion strategy. In particular, the predictions provided by the long short‐term memory network are less affected by erroneous assumptions on the noise statistics and prior model and by errors in the estimated source wavelet.

show abstract

“…This presumption, however, makes the network focus mainly on memorizing a large amount of input information and weakens its modeling capability [35][36]. To make up for this deficiency, skip connections [37][38][39], the core technique of DRN [40], is introduced especially for deeper Bi-LSTM networks, because each neuron node in the skip connections makes uses of the information of previous hidden layer and enhances the modeling ability of network.…”

Section: Introductionmentioning

confidence: 99%

Attention-Based Convolution Skip Bidirectional Long Short-Term Memory Network for Speech Emotion Recognition

2021

View full text Add to dashboard Cite

Seismic Data Reconstruction Using Deep Bidirectional Long Short-Term Memory With Skip Connections

Cited by 56 publications

References 17 publications

Revisit Geophysical Imaging in A New View of Physics-informed Generative Adversarial Learning

Revisit Geophysical Imaging in A New View of Physics-informed Generative Adversarial Learning

Elastic properties and litho‐fluid facies estimation from pre‐stack seismic data through bidirectional long short‐term memory networks

Attention-Based Convolution Skip Bidirectional Long Short-Term Memory Network for Speech Emotion Recognition

Contact Info

Product

Resources

About