AtomSegNet by DeepEM Lab@UCI

Please reference paper: Scientific Reports, 11, 5386 (2021) https://doi.org/10.1038/s41598-021-84499-w

pytorch == 1.3.0 torchvision==0.4.0 scikit-image=0.18.3 (up to 0.18.3. Incompatible with 0.19 because watershed has been moved to a different module)

Try the following if you have trouble loading the app

conda install scikit-image=0.18.3

pip install scikit-image==0.18.3

python Code_Atom_Seg_Ui.py

In DOS terminal, enter

run

We beat the golden standard--2d Gaussian fit

See more examples at Reports, 11, 5386 (2021) https://doi.org/10.1038/s41598-021-84499-w