NeuralMetamaterialNetwork
Cpp and python code for our SIGGRPAH Asia 2023 paper Neural Metamaterial Network for Nonlinear Material Design
Install / Use
/learn @liyuesolo/NeuralMetamaterialNetworkREADME
Neural Metamaterial Networks for Nonlinear Material Design

Yue Li, Stelian Coros, Bernhard Thomaszewski
ACM Transactions on Graphics (Proc. SIGGRAPH Asia 2023)
Paper Video
Code Structure
Projects/NeuralMetamaterial/
- [include & src] contains header and source files for the FEM solver
- [python] contains pretained models, code for training and inverse design
- [scripts] for generating training data
- [data] example data file for training
Compile and Run
Simply execute the build.sh file
Third Party Library Required
The version that is used to compile this code under Ubuntu 20.04 is attached as ThirdParty.zip
Libigl (https://libigl.github.io/)
Gmsh (https://gmsh.info/)
IPC Toolkit (https://ipc-sim.github.io/ipc-toolkit/)
SuiteSparse (https://github.com/DrTimothyAldenDavis/SuiteSparse)
These folder should be in the same folder as the base folder.
The versions that were used for this repo can be found here.
Third Party Attached
Tactile (https://github.com/isohedral/tactile)
Clipper (https://github.com/AngusJohnson/Clipper2)
SpectrA (https://spectralib.org/)
Poisson Disk Sampling (http://www.cemyuksel.com/cyCodeBase/soln/poisson_disk_sampling.html)
License
See the LICENSE file for license rights and limitations.
Contact
Please reach out to me (Yue Li) if you spot any bugs or having quesitons or suggestions!
