🚧 Vimentin
easymode segment vault --2d
The current vimentin model is a 3D Ais UNet trained on single annotated slices from different human cell lines including Jurkat, HeLa, fibroblasts, and neutrophils. It has not been validated by subtomogram averaging and we currently only have a 2.5D network.
Example output
Example of easymode segment vimentin --2d output overlaid on a tomogram of a human T lymphocyte (Jurkat).