Reproducible material for Multiple Wavefield Solutions in Physics-Informed Neural Networks using Latent Representation - Mohammad H. Taufik, Xinquan Huang, Tariq Alkhalifah.

Project structure

This repository is organized as follows:

• 📂 asset: folder containing logo.
• 📂 data: a folder containing the subsampled velocity models used to train the PINN.
• 📂 notebooks: reproducible notebook for the synthetic tests of the paper.
• 📂 scripts: script examples to perform autoencoder training, PINNs training to solve for the eikonal and scattered Helmholtz equations.
• 📂 saves: a folder containing the trained PINN model.
• 📂 src: a folder containing routines for the latentpinn source file.

Getting started 👾 🤖

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

To install the environment, run the following command:

./install_env.sh

It will take some time, but if, in the end, you see the word Done! on your terminal, you are ready to go.

Remember to always activate the environment by typing:

conda activate latentpinn

Disclaimer: All experiments have been carried on a Intel(R) Xeon(R) Silver 4316 CPU @ 2.30GHz equipped with a single NVIDIA A100 GPU. Different environment configurations may be required for different combinations of workstation and GPU.

Cite us

@article{taufik2023latentpinns,
  title={LatentPINNs: Generative physics-informed neural networks via a latent representation learning},
  author={Taufik, Mohammad H and Alkhalifah, Tariq},
  journal={arXiv preprint arXiv:2305.07671},
  year={2023}
}
@article{taufik2024multiple,
  title={Multiple Wavefield Solutions in Physics-Informed Neural Networks using Latent Representation},
  author={Taufik, Mohammad H and Huang, Xinquan and Alkhalifah, Tariq},
  journal={IEEE Geoscience and Remote Sensing Letters},
  year={2024},
  publisher={IEEE}
}