The repository contains codes for our work titled "Joint learning for seismic inversion: An acoustic impedance estimation case study" accepted presented at the SEG 2020 Annual Meeting and published to SEG 2020 Expanded Abstracts. The preprint for the work may be obtained here.
The complete bibtex citation for the work is as follows:
@inbook{doi:10.1190/segam2020-3428378.1,
author = {Ahmad Mustafa and Ghassan AlRegib},
title = {Joint learning for seismic inversion: An acoustic impedance estimation case study},
booktitle = {SEG Technical Program Expanded Abstracts 2020},
chapter = {},
pages = {1686-1690},
year = {2020},
doi = {10.1190/segam2020-3428378.1},
URL = {https://library.seg.org/doi/abs/10.1190/segam2020-3428378.1},
}
The repository contains all the data used to run the codes. Clone the github repo to an appropriate directory on your machine. You may use a dedicated IDE like Pycharm or Spyder to run the main.py file, which will extract the data, perform training, and then use the trained models to infer on the Marmousi 2 and SEAM sections to print out the estimated 2-D profiles as well as various regression metrics between the estimated and ground-truth impedance for both datasets. Alternatively, you may use the command line to to run codes and print results, as follows:
cd <project root directory>
python main.py --no_wells_marmousi 50 --no_wells_seam 12 --epochs 900 --gamma 0.0001
In case you face problems with running the codes, please contact Ahmad Mustafa at amustafa9@gatech.edu.