Realistically Textured Random Velocity Models for Deep Learning Applications

Kazei, Vladimir; Ovcharenko, Oleg; Alkhalifah, Tariq; Simons, F.H.

doi:10.3997/2214-4609.201901340

Cited by 13 publications

(8 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Then we replicate the produced log of velocity perturbations for desired width of the model and add the 1D seed trend from field data. Next, we follow Kazei et al (2019a) and apply an elastic transform to distort the layered model. Finally, we select the maximum value between the 1D seed trend and the produced model (Figure 1).…”

Section: Generation Of the Training Datasetmentioning

confidence: 99%

Transferring Elastic Low Frequency Extrapolation from Synthetic to Field Data

Ovcharenko¹,

Kazei

Peter³

et al. 2021

82nd EAGE Annual Conference &Amp; Exhibition

Self Cite

View full text Add to dashboard Cite

Training deep learning models on synthetic data is a common practice in geophysics. However, knowledge transfer from synthetic to field applications is often a bottleneck. Here, we describe the workflow for generation of realistic synthetic dataset of elastic waveforms, sufficient for low-frequency extrapolation in marine streamer setup. Namely, we first extract the source signature, the noise imprint and a 1D velocity model from real marine data. Then, we use those to generate pseudo-random initializations of elastic subsurface models and simulate elastic wavefield data. After that, we enrich the simulated data with realistic noise and use it to train a deep neural network. Finally, we demonstrate the results of low-frequency extrapolation on field streamer data, given that the model was trained exclusively on a synthetic dataset.

show abstract

Section: Generation Of the Training Datasetmentioning

confidence: 99%

Transferring Elastic Low Frequency Extrapolation from Synthetic to Field Data

Ovcharenko¹,

Kazei

Peter³

et al. 2021

82nd EAGE Annual Conference &Amp; Exhibition

Self Cite

View full text Add to dashboard Cite

show abstract

“…In the age of digital transformation, deep learning models have been used widely in many seismic applications such as data processing (Ovcharenko et al, 2019;Kazei et al, 2019), modeling (Song et al, 2021), inversion (Araya-Polo et al, 2018;Kazei et al, 2020;Sun and Alkhalifah, 2020) and interpretation (Xiong et al, 2018;Sen et al, 2020). The type of neural network architecture needed depends on the application.…”

Section: Introductionmentioning

confidence: 99%

Time-lapse data matching using a recurrent neural network approach

et al. 2022

View full text Add to dashboard Cite

Time-lapse seismic data acquisition is an essential tool to monitor changes in a reservoir due to fluid injection, such as CO2 injection. By acquiring multiple seismic surveys in the exact same location, we can identify the reservoir changes by analyzing the difference in the data. However, such analysis can be skewed by the near-surface seasonal velocity variations, inaccuracy and repeatability in the acquisition parameters, and other inevitable noise. The common practice (cross-equalization) to address this problem uses the part of the data where changes are not expected to design a matching filter and then apply it to the whole data, including the reservoir area. Like cross-equalization, we train a recurrent neural network on parts of the data excluding the reservoir area and then infer the reservoir-related data. The recurrent neural network can learn the time dependency of the data, unlike the matching filter that processes the data based on the local information obtained in the filter window. We demonstrate the method of matching the data in various examples and compare it with the conventional matching filter. Specifically, we start by demonstrating the ability of the approach in matching two traces and then test the method on a pre-stack 2D synthetic data. Then, we verify the enhancements of the 4D signal by providing RTM images. We measure the repeatability using normalized root mean square and predictability metrics and show that in some cases, our proposed method performed better than the matching filter approach.

show abstract

“…However, the datasets should be statistically similar. Despite a number of approaches proposed for generation of realistic seismic data (Kazei et al, 2019), the knowledge transfer between synthetic and field applications remains challenging.…”

Section: Introductionmentioning

confidence: 99%

Dual-band generative learning for low-frequency extrapolation in seismic land data

Ovcharenko

Kazei

Peter

et al. 2021

First International Meeting for Applied Geoscience &Amp; Energy Expanded Abstracts

View full text Add to dashboard Cite

The presence of low-frequency energy in seismic data is one of the desirable prerequisite for a successful full-waveform inversion. However, it requires costly machinery and operations to acquire such data. Extrapolation for the missing lowfrequency content in field data might be addressed in a datadriven framework. In particular, the deep learning models are commonly trained for this purpose on synthetic data and then used for inference on field data. However, it is always a challenge to switch the application domains. We propose the concept of generative dual-band learning to facilitate the knowledge transfer between synthetic and field seismic data applications of low-frequency data extrapolation. We first explain the two-step procedure for training of a generative adversarial network (GAN) that extrapolates low frequencies. Then we describe the workflow for synthetic dataset generation. Finally, we explore the feasibility of the dual-band learning concept on real near-surface land data acquired in Saudi Arabia.

show abstract

Realistically Textured Random Velocity Models for Deep Learning Applications

Cited by 13 publications

References 5 publications

Transferring Elastic Low Frequency Extrapolation from Synthetic to Field Data

Transferring Elastic Low Frequency Extrapolation from Synthetic to Field Data

Time-lapse data matching using a recurrent neural network approach

Dual-band generative learning for low-frequency extrapolation in seismic land data

Contact Info

Product

Resources

About