StorSeismic: An approach to pre-train a neural network to store seismic data features

This repository contains codes and resources to reproduce experiments of StorSeismic in Harsuko and Alkhalifah, 2020.

Requirements

We use RAdam as the default optimizer. To install this, use:

pip install git+https://github.com/LiyuanLucasLiu/RAdam

Instruction

No	Notebook name	Description
1	nb0_1_data_prep_pretrain.ipynb	Create pre-training data
2	nb0_2_data_prep_finetune.ipynb	Create fine-tuning data
3	nb1_pretraining.ipynb	Pre-training of StorSeismic
4	nb2_1_finetuning_denoising.ipynb	Example of fine-tuning task: denoising
5	nb2_2_finetuning_velpred.ipynb	Example of fine-tuning task: velocity estimation

References

Harsuko, R., & Alkhalifah, T. A. (2022). StorSeismic: A new paradigm in deep learning for seismic processing. IEEE Transactions on Geoscience and Remote Sensing, 60, 1-15.

Citation

Citations are very welcomed. This work can be cited using:

@article{harsuko2022storseismic,
  title={StorSeismic: A new paradigm in deep learning for seismic processing},
  author={Harsuko, Randy and Alkhalifah, Tariq A},
  journal={IEEE Transactions on Geoscience and Remote Sensing},
  volume={60},
  pages={1--15},
  year={2022},
  publisher={IEEE}
}