napari-easy-augment-batch-dl

See full documentation

A plugin to perform deep learning on small to medium sized image sets with UNETs, Cellpose, Stardist, SAM and friends. In particular this plugin is useful for performing deep learning with a small number of labels and augmentation, and experimenting with different deep learning frameworks.

Important note on dependencies: This plugin is designed to work with different permutations of dependencies. For example it should work if one of Pytorch, Cellpose, SAM and/or Stardist is installed but does not require all. Thus we don't specify all the dependencies and leave it up to the user to install the permutation of DL related dependencies they would like to use. More detailed instructions are below.

If you have any questions about dependencies splease post on the Image.sc forum.

This napari plugin was generated with Cookiecutter using @napari's cookiecutter-napari-plugin template.

Installation

To install latest development version :

pip install git+https://github.com/bnorthan/napari-easy-augment-batch-dl.git

You will also need to install the latest development version of tnia-python:

pip install git+https://github.com/True-North-Intelligent-Algorithms/tnia-python.git

You will need to install napari and for augmentation you will need albumentations library. Also explicitly install numpy 1.26. (We have not tested with numpy 2.0 so it is a good idea to explicitly install numpy 1.26 to avoid another dependency installing numpy 2.x)

    pip install numpy==1.26
    pip install napari[all]
    pip install albumentations
    pip install matplotlib

You will also need one or more of stardist, cellpose, segment-everything or Yolo

Stardist

Windows

    conda install -c conda-forge cudatoolkit=11.2 cudnn=8.1.0
    pip install "tensorflow<2.11"
    pip install stardist==0.8.5
    pip install gputools
    pip install edt

Linux

    pip install tensorflow[and-cuda]
    pip install stardist
    pip install gputools
    pip install edt

Pytorch (for unet segmentation)

    pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118
    pip install pytorch-lightning
    pip install monai
    pip install scipy
    pip install tifffile

Cellpose

    pip install cellpose

SAM (Segment Anything)

    pip install segment-everything

You can install napari-easy-augment-batch-dl via pip:

pip install napari-easy-augment-batch-dl

Contributing

Contributions are very welcome. Tests can be run with tox, please ensure the coverage at least stays the same before you submit a pull request.

License

Distributed under the terms of the BSD-3 license, "napari-easy-augment-batch-dl" is free and open source software

Issues

If you encounter any problems, please file an issue along with a detailed description.