Skip to content

Groupwise Combinatorial Multi-Branch U-Net for atomic segmentation of moiré lattices

Notifications You must be signed in to change notification settings

ahoust17/Gomb-Net

Repository files navigation

Gomb-Net

Revealing atomic identities in bilayer moiré materials

2024O Oct 22

Gomb-Net finds atoms, layer-wise, in STEM images of moiré materials

Gomb-Net is a multi-branch U-Net which trains using a unique groupwise output handling and combinatorial loss function.

Want to use Gomb-Net, but unsure how to start?

Reach out!

Austin Houston

ahoust17@vols.utk.edu

Try running the following notebooks on google colab (button in the notebooks):

Eval_Graphene_model.ipynb

Eval_WSSe_model.ipynb

all experimental data is available through:

https://drive.google.com/file/d/1DyKtrmJ8wNYQg3YEJ8_iXjz6lB_DQfwy/view?usp=sharing

About

Groupwise Combinatorial Multi-Branch U-Net for atomic segmentation of moiré lattices

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published