UCSB-CASL / Curvature_ECNet_2D

Error-correcting neural networks for two-dimensional curvature computation in the level-set method

Error-correcting neural networks for two-dimensional curvature computation in the level-set method

This is the accompanying repository for our manuscript.

Contents

In addition to the usual requirements.txt, we include the following files under the corresponding resolution folder within models/:

1. fdeep_k_nnet.json: Neural network as exported by the frugally-deep library (without optimizer).
2. k_nnet.h5: Tensorflow/Keras model in HDF5 format (without optimizer).
3. k_nnet.json: Our custom JSON version of the neural network with hidden-layer weights encoded in Base64 but decoded as ASCII text. The "output" key refers to the last hidden layer. It does not include the aditive neuron.
4. k_pca_scaler.pkl: PCA scaler stored in pickle format.
5. k_pca_scaler.json: JSON version of PCA scaler with plain-valued parameters.
6. k_std_scaler.pkl: Standard scaler in pickle format.
7. k_std_scaler.json: JSON version of standard scaler with plain-valued parameters.

Testing data sets

The flowers directory contains the data sets we used for testing our curvature hybrid solvers. The corresponding input files contain samples that have been reoriented per the description provided in the manuscript. It is still necessary, though, that each sample be reflectect to take the average of two network evaluations.

Contact

Please forward your questions to this email.