amirmardan / pinn_fwi

PGNN-FWI: performing physics-guided neural network for FWI

Hi there 👋

I'm Amir, a postdoctoral researcher at Polytechnique Montréal working on applications of AI for processing and inversion of near-surface seismic data. I'm based in Quebec City, Canada.

🔭 I’m currently working on

• processing near-surface seismic data using AI,
• velocity modelling using deep learning,
• velocity modelling using FWI.

📚 These days, I'm mostly reading about

• self-supervised learning,
• uncertainty analysis and decision-making under uncertainty.

You can see my research at my personal page

📫 How to reach me: [email protected]

My main public repositories

Repository	Description
PyFWI	A Python package for performing FWI and TL-FWI
pinn_fwi	An implementation of PINN for performing FWI
ml_course	Prepared materials for my machine learning course at AUT
first_break_picking	A Python package for performing automatic first-break picking

⚡ Technologies

pinn_fwi's Issues

Reproducibility

The result is not reproducible when we use multiscale fwi, seems there is a bug in lpass
https://github.com/AmirMardan/pinn_fwi/blob/main/torchfwi.py#L31-L33