Reproducible material for Conditional Image Prior for Uncertainty Quantification in Full Waveform Inversion-Lingyun Yang, Omar M. Saad, Guochen wu and Tariq Alkhalifah
This repository is organized as follows:
- 📂 data: folder containing data (or instructions on how to retrieve the data;
- 📂 Result: folder containing marmousi model FWI results;
The following notebooks are provided:
- 📙
Marmousi_ContestUnet.ipynb: Pretraining stage for the conditional CNN using 50 particel; - 📙
Marmousi_FWI-part.ipynb: Performing UQ FWI using the pre-trained conditional CNN.
To ensure reproducibility of the results, we suggest using the following pip environment.
Simply run:
conda create -n deepwaveold_env python=3.7
conda activate deepwaveold_env
pip install torch==1.7.1+cu110 torchvision==0.8.2+cu110 torchaudio==0.7.2 -f https://download.pytorch.org/whl/torch_stable.html
pip install deepwave==0.0.8
pip install gstools
pip install torchsummary
pip install scikit-learn
pip install pylops
pip install hydra-core
It will take some time, if at the end you see the word Done! on your terminal you are ready to go.
Disclaimer: All experiments have been carried on a Intel(R) Xeon(R) CPU @ 2.10GHz equipped with a single NVIDIA GEForce RTX 3090 GPU. Different environment configurations may be required for different combinations of workstation and GPU.
Lingyun Yang, Omar M. Saad, Tariq Alkhalifah, et al. 2026. Conditional image prior for uncertainty quantification in full waveform inversion.