Neural networks for GPR radargram inversion, trained on synthetic gprMax data
-
Updated
Oct 3, 2026 - Jupyter Notebook
Neural networks for GPR radargram inversion, trained on synthetic gprMax data
Runs gprMax simulations by generating input models, applied to glaciology.
. It is a powerful simulation software designed to generate received signals from a variety of simulated scenarios, allowing for the labeling and saving of the resulting data.
Google Summer of Code 2025 Demo Implementation : AI Chatbot for support
Automated integration testing and physics-based fidelity validation for gprMax using NRMSE regression gates and CI/CD pipelines.
To associate your repository with the gprmax topic, visit your repo's landing page and select "manage topics."