elektronn 1.0.14

A highly configurable toolkit for training 3d/2d CNNs and general Neural Networks

ELEKTRONN is a highly configurable toolkit for training 3D/2D CNNs and general Neural Networks.

It is written in Python 2 and based on Theano, which allows CUDA-enabled GPUs to significantly accelerate the pipeline.

The package includes a sophisticated training pipeline designed for classification/localisation tasks on 3D/2D images. Additionally, the toolkit offers training routines for tasks on non-image data.

ELEKTRONN was created by Marius Killinger and Gregor Urban at the Max Planck Institute For Medical Research to solve connectomics tasks.

Membrane and mitochondria probability maps. Predicted with a CNN with recursive training. Data: zebra finch area X dataset j0126 by Jörgen Kornfeld.

Learn More:

Website

Installation instructions

Documentation

Source code

Toy Example

$ elektronn-train MNIST_CNN_warp_config.py

This will download the MNIST data set and run a training defined in an example config file. The plots are saved to ~/CNN_Training/2D/MNIST_example_warp.

File structure

ELEKTRONN
├── doc                     # Documentation source files
├── elektronn
│   ├── examples            # Example scripts and config files
│   ├── net                 #  Neural network library code
│   ├── scripts             #  Training script and profiling script
│   ├── training            #  Training library code
│   └── ...
├── LICENSE.rst
├── README.rst
└── ...