Deep learning seismic facies on state-of-the-art CNN architectures

Jesper S. Dramsch, Technical University of Denmark, and Mikael Lüthje, Technical University of Denmark

Abstract

We explore propagation of seismic interpretation by deep learning in stacked 2D sections. We show the application of state-of-the-art image classification algorithms on seismic data. These algorithms were trained on big labeled photograph databases. We use transfer learning to benefit from pre-trained networks and evaluate their performance on seismic data.

Presentation Date: Wednesday, October 17, 2018
Start Time: 8:30:00 AM
Location: 204B (Anaheim Convention Center)
Presentation Type: Oral

Citation

Paper

Jesper S. Dramsch and Mikael Lüthje (2018) Deep-learning seismic facies on state-of-the-art CNN architectures. SEG Technical Program Expanded Abstracts 2018: pp. 2036-2040.

Presentation

Dramsch, Jesper Soeren; Lüthje, Mikael (2018): Deep-learning seismic facies on state-of-the-art CNN architectures. figshare. Presentation. https://doi.org/10.6084/m9.figshare.7301645.v1

Code

Dramsch, Jesper Soeren; Lüthje, Mikael (2018): Deep-learning seismic facies on state-of-the-art CNN architectures. figshare. Code. https://doi.org/10.6084/m9.figshare.7227545

Usage

• Open the Notebook
• Download the F3 Seismic Data
• Download Models from the Model Zoo
• Have Fun Experimenting

Interpretation of VGG

Loss of VGG

References

• Abadi, M., A. Agarwal, P. Barham, E. Brevdo, Z. Chen, C. Citro, G. S. Corrado, A. Davis, J. Dean, M. Devin, S. Ghemawat, I. Goodfellow, A. Harp, G. Irving, M. Isard, Y. Jia, R. Jozefowicz, L. Kaiser, M. Kudlur, J. Levenberg, D. Mane, R. Monga, S. Moore, D. Murray, C. Olah, M. Schuster, J. Shlens, B. Steiner, I. Sutskever, K. Talwar, P. Tucker, V. Vanhoucke, V. Vasudevan, F. Viegas, O. Vinyals, P. Warden, M. Wattenberg, M. Wicke, Y. Yu, and X. Zheng, 2015, TensorFlow: Large-scale machine learning on heterogeneous systems. (Software available from tensorflow.org).
• Baxter, J., 1998, Theoretical models of learning to learn, in Learning to learn: Springer, 71–94.
• Charles Rutherford Ildstad, P. B., 2017, MalenoV. Machine learning of Voxels.
• Chollet, F., et al., 2015, Keras,
• Dahl, G. E., T. N. Sainath, and G. E. Hinton, 2013, Improving deep neural networks for LVCSR using rectified linear units and dropout: Presented at the IEEE International Conference on Acoustics Speech and Signal Processing.
• Deng, J., W. Dong, R. Socher, L.-J. Li, K. Li, and L. Fei-Fei, 2009, ImageNet: A large-scale hierarchical image database: Presented at the CVPR09.
• He, K., X. Zhang, S. Ren, and J. Sun, 2016, Deep residual learning for image recognition: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 770–778.
• Krizhevsky, A., I. Sutskever, and G. E. Hinton, 2012, ImageNet classification with deep convolutional neural networks, in Advances in neural information processing systems: Curran Associates, Inc. 25, 1097–1105.
• Lecun, Y., 1989, Generalization and network design strategies, in Connectionism in perspective: Elsevier.
• Lin, T.-Y., P. Goyal, R. Girshick, K. He, and P. Dollar, 2017, Focal loss for dense object detection: arXiv preprint arXiv:1708.02002.
• Long, J., E. Shelhamer, and T. Darrell, 2015, Fully convolutional networks for semantic segmentation: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 3431–3440.
• Ruder, S., 2016, An overview of gradient descent optimization algorithms: arXiv preprint arXiv:1609.04747.
• Rumelhart, D., G. Hinton, and R. Williams, 1988, Learning internal representations by error propagation, in Readings in cognitive science: Elsevier, 399–421.
• Simonyan, K., and A. Zisserman, 2014, Very deep convolutional networks for large-scale image recognition: arXiv preprint arXiv:1409.1556.
• Srivastava, N., G. Hinton, A. Krizhevsky, I. Sutskever, and R. Salakhutdinov, 2014, Dropout: A simple way to prevent neural networks from overfitting: Journal of Machine Learning Research, 15, 1929–1958.
• Waldeland, A., and A. Solberg, 2016, 3D attributes and classification of salt bodies on unlabelled datasets: 78th Annual International Conference and Exhibition, EAGE, Extended Abstracts, https://doi.org/10.3997/2214-4609.201600880
• Widrow, B., and M. Lehr, 1990, 30 years of adaptive neural networks: Perceptron Madaline, and backpropagation: Proceedings of the IEEE, 78, 1415–1442, https://doi.org/10.1109/5.58323
• Yilmaz, Ö., 2001, Seismic data analysis: SEG.

Notes

We explore transfer training for automatic seismic interpretation without fine-tuning. See and cite the Powerpoint

Read More: https://library.seg.org/doi/abs/10.1190/segam2018-2996783.1 Or at: https://dramsch.net/#portfolio