1887

Abstract

Summary

Stacking velocity analysis is conventionally performed by manual picking on semblance spectra, a task that is both labour intensive and time consuming. While recent advancements in machine learning have enabled automated approaches, significant challenges remain, especially when dealing with complex data scenarios. We propose a novel approach for automated stacking velocity analysis using convolutional long short-term memory (ConvLSTM) networks, combining the feature extraction capabilities of convolutional neural networks (CNNs) with the sequence modeling strengths of long short-term memory (LSTM) networks. The model is trained on synthetic seismic data with free surface multiples and generated from structurally diverse velocity models and tested on complex datasets, including the Marmousi model. We further demonstrate the advantages offered by transfer learning in refining the performance of the model on unseen and complex test data. In particular, transfer learning requires relatively limited amount of data retraining and less compute time, owing to the simplicity of the network architecture. ConvLSTM networks can constitute efficient and accurate tools for automated stacking velocity analysis in seismic data processing.

© EAGE Publications BV
Loading

Article metrics loading...

/content/papers/10.3997/2214-4609.202539042
2025-03-24
2026-02-06

  • From This Site

    /content/papers/10.3997/2214-4609.202539042
    dcterms_title,dcterms_subject,pub_keyword
    -contentType:Journal -contentType:Contributor -contentType:Concept -contentType:Institution
    10
    5
Loading full text...

Full text loading...

References

  1. Fabien-Ouellet, G. and Sarkar, R. [2020] Seismic velocity estimation: A deep recurrent neural-network approach. Geophysics, 85(1), U21–U29.
     [Google Scholar]
  2. Fomel, S. [2009] Velocity analysis using AB semblance. Geophysical Prospecting, 57(3), 311–321.
     [Google Scholar]
  3. Foster, D. [2019] Generative Deep Learning - Teaching Machines to Paint, Write, Compose and Play. O'Reilly.
     [Google Scholar]
  4. Hochreiter, S. and Schmidhuber, J. [1997] Long Short-Term Memory. Neural Computation, 9(8), 1735–1780.
     [Google Scholar]
  5. Kingma, D.P. and Ba, J. [2015] Adam: A Method for Stochastic Optimization. In: 3rd International Conference on Learning Representations (ICLR).
     [Google Scholar]
  6. Park, M.J. and Sacchi, M.D. [2020] Automatic velocity analysis using convolutional neural network and transfer learning. Geophysics, 85(1), V33–V43.
     [Google Scholar]
  7. Richardson, A. [2023] Deepwave. doi:10.5281/zenodo.8381177.
     https://doi.org/10.5281/zenodo.8381177 [Google Scholar]
  8. Shi, X., Chen, Z., Wang, H., Yeung, D.Y., kin Wong, W. and chun Woo, W. [2015] Convolutional LSTM Network: A Machine Learning Approach for Precipitation Nowcasting.
     [Google Scholar]
  9. Simonyan, K. and Zisserman, A. [2015] Very Deep Convolutional Networks for Large-Scale Image Recognition.
     [Google Scholar]
  10. Toldi, J.L. [1989] Velocity analysis without picking. Geophysics, 54(2), 191–199.
     [Google Scholar]
  11. Versteeg, R. [1994] The Marmousi experience: Velocity model determination on a synthetic complex data set. The Leading Edge, 13(9), 927–936.
     [Google Scholar]
  12. bin Waheed, U., Al-Zahrani, S. and Hanafy, S.M. [2019] Machine learning algorithms for automatic velocity picking: K-means vs. DBSCAN. In: SEG International Exposition and Annual Meeting. SEG, 5110–5114.
     [Google Scholar]
/content/papers/10.3997/2214-4609.202539042
Loading
Automated Stacking Velocity Analysis using ConvLSTM Networks and Transfer Learning
Conference Proceedings 2025, 1 (2025); https://doi.org/10.3997/2214-4609.202539042
/content/papers/10.3997/2214-4609.202539042
/content/papers/10.3997/2214-4609.202539042
Loading

Data & Media loading...

Most Cited This Month Most Cited RSS feed

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