Reproducible material for A self-supervised learning framework for seismic low-frequency extrapolation - Shijun Cheng, Yi Wang, Qingchen Zhang, Randy Harsuko, Tariq Alkhalifah.

Project structure

This repository is organized as follows:

• 📂 ssl_lowfreq: python library containing routines for self-supervised low-frequency extrapolation;
• 📂 asset: folder containing logo;
• 📂 data: folder to store dataset;
• 📂 scripts: set of shell scripts used to run training and testing

Supplementary files

To ensure reproducibility, we provide the the data set for training and testing, and the trained model for validate the low-frequency extrapolation performance.

• Training and Testing data set Download the training and testing data set here. Then, use unzip to extract the contents to dataset/train and dataset/test.

• Trained model Download the trained neural network model here. Then, extract the contents to trained_checkpoints/.

Getting started 👾 🤖

To ensure reproducibility of the results, we suggest using the environment.yml file when creating an environment.

Simply run:

./install_env.sh

It will take some time, if at the end you see the word Done! on your terminal you are ready to go. Activate the environment by typing:

conda activate  ssl_lowfreq

After that you can simply install your package:

pip install .

or in developer mode:

pip install -e .

Configuration file for training and testing:

The training and testing configuration is organized in the config file config.yaml. You can modify the configuration value according to your situation.

Scripts 📄

When you have downloaded the supplementary files and have installed the environment, you can entry the scripts file folder and run demo. We provide two scripts which are responsible for training and testing examples.

Under our framework, for training, you need to set configuration value of args (train_mode) in config file config.yaml as SSL, and then directly run:

sh run_train.sh

If you need to compare with a supervised learning framework, you need to set configuration value of args (train_mode) in config file config.yaml as SL, and then run:

sh run_train.sh

For testing, you can directly run:

sh run_test.sh

Note: Here, we have provided a trained model in supplementary file, you can directly load trained model to perform testing.

Note: We emphasize that the training logs is saved in the runs/ file folder. You can use the tensorboard --logdir=./ or extract the log to view the changes of the metrics as a function of epoch.

Disclaimer: All experiments have been carried on a Intel(R) Xeon(R) CPU @ 2.10GHz equipped with a single NVIDIA GEForce A100 GPU. Different environment configurations may be required for different combinations of workstation and GPU. Due to the high memory consumption during the training phase, if your graphics card does not support large batch training, please reduce the configuration value of args (batch_size) in config file config.yaml.

Cite us

@article{cheng2024self,
  title={A Self-Supervised Learning Framework for Seismic Low-Frequency Extrapolation},
  author={Cheng, Shijun and Wang, Yi and Zhang, Qingchen and Harsuko, Randy and Alkhalifah, Tariq},
  journal={Journal of Geophysical Research: Machine Learning and Computation},
  volume={1},
  number={3},
  pages={e2024JH000157},
  year={2024},
  publisher={Wiley Online Library}
}