Skip to content

Contains trained cGANs, data, and MATLAB scripts used in our recent paper

Notifications You must be signed in to change notification settings

nuwan-d/deep_generative_neural_net

Repository files navigation

Uncovering Stress Fields and Defect Distributions in Graphene Using Deep Neural Networks

This repository contains the MATLAB scripts used to generate LAMMPS data/input files as well as postprocessing the numerical results of the simulations (i.e., to transformed into image-based data) in our recent article in the International Journal of Fracture (2023). A video abstract of the article is available here.

The architecture of the cGAN is shown in Fig. 3a. The generator is trained to produce images that the discriminator cannot distinguish from real images. The discriminator is adversarially trained to distinguish between predicted images (Fig. a) and real images (Fig. b).

The trained neural networks and the complete data set are freely available here.

About

Contains trained cGANs, data, and MATLAB scripts used in our recent paper

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages