Example

To train the ICNN to learn the hidden Isihara model from the dataset with high noise conditioning run the following:

• python main.py Isihara high

The script will train a number of neural networks (specified in config.py) and evaluate each model against the ground-truth model along six deformation paths. The models will be saved in results/Isihara/. A plot will be saved in results/ including strain energy density and Piola Kirchhoff stress tensor component evaluations along the deformation paths.