Train model for nucleus segmentation

[constantinpape]

The following datasets could be relevant:

• https://github.com/constantinpape/torch-em/blob/main/torch_em/data/datasets/light_microscopy/embedseg_data.py
• https://github.com/constantinpape/torch-em/blob/main/torch_em/data/datasets/light_microscopy/gonuclear.py
• https://github.com/constantinpape/torch-em/blob/main/torch_em/data/datasets/light_microscopy/mouse_embryo.py
• https://github.com/constantinpape/torch-em/blob/main/torch_em/data/datasets/light_microscopy/plantseg.py
• https://github.com/constantinpape/torch-em/blob/main/torch_em/data/datasets/light_microscopy/blastospim.py

@anwai98 can also provide more context on them.

(We want to train a model for nucleus segmentation, based o DAPI and/or Hoechst in light sheet micrsocopy)

[constantinpape]

For each dataset you can use get_..._data to download the respective dataset.

[anwai98]

There's some more data for this:

• NIS3d (light-sheet imaged data) with annotated nuclei: https://github.com/constantinpape/torch-em/blob/main/torch_em/data/datasets/light_microscopy/nis3d.py
• Data from Loic's lab (a lot of light sheet data, but not implemented yet in torch-em - if we'd like more data, I can prepare it super quickly, the zarr files are open-source, Jordao pointed the open ones to me) https://www.biorxiv.org/content/10.1101/2024.09.02.610652v1

(will add more in this comment if I find any around ;))