https://github.com/funkelab/gunpowder

A library to facilitate machine learning on multi-dimensional images.
https://github.com/funkelab/gunpowder

computer-vision gunpowder machine-learning multidimensional pipeline

Last synced: 6 days ago
JSON representation

A library to facilitate machine learning on multi-dimensional images.

• Host: GitHub
• URL: https://github.com/funkelab/gunpowder
• Owner: funkelab
• License: mit
• Created: 2017-04-11T15:54:43.000Z (over 7 years ago)
• Default Branch: main
• Last Pushed: 2024-09-25T19:18:55.000Z (about 2 months ago)
• Last Synced: 2024-11-08T11:52:16.592Z (7 days ago)
• Topics: computer-vision, gunpowder, machine-learning, multidimensional, pipeline
• Language: Python
• Homepage: https://funkelab.github.io/gunpowder/
• Size: 30.1 MB
• Stars: 77
• Watchers: 13
• Forks: 56
• Open Issues: 10
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

• awesome-janelia-software - Gunpowder - Library for machine learning on multi-dimensional images (Machine Learning)
• awesome-janelia-software - Gunpowder - Library for machine learning on multi-dimensional images (Machine Learning)

README

        
![gunpowder](docs/source/img/gunpowder.svg)

[![Tests](https://github.com/funkey/gunpowder/actions/workflows/test.yml/badge.svg)](https://github.com/funkey/gunpowder/actions/workflows/test.yml)

A library to facilitate machine learning on large, multi-dimensional images.

`gunpowder` allows you to assemble a pipeline from

[data loading](http://funkelab.github.io/gunpowder/api.html#source-nodes)

over

[pre-processing](http://funkelab.github.io/gunpowder/api.html#image-processing-nodes),

[random batch sampling](http://funkelab.github.io/gunpowder/api.html#randomlocation),

[data augmentation](http://funkelab.github.io/gunpowder/api.html#augmentation-nodes),

[pre-caching](http://funkelab.github.io/gunpowder/api.html#precache),

[training/prediction](http://funkelab.github.io/gunpowder/api.html#training-and-prediction-nodes), to

[storage of results](http://funkelab.github.io/gunpowder/api.html#output-nodes)

on arbitrarily large volumes of

multi-dimensional images. `gunpowder` is not tied to a particular learning

framework, and thus complements libraries like

[`torch`](https://pytorch.org/),

[`tensorflow`](https://www.tensorflow.org/).

The full documentation can be found [here](https://funkelab.github.io/gunpowder).

`gunpowder` was originally written by Jan Funke and is inspired by

[`PyGreentea`](https://github.com/TuragaLab/PyGreentea) by William Grisaitis,

Fabian Tschopp, and Srini Turaga.