Citation

If you use any of this code in a scientific publication, please cite:

Schirrmeister, R. T., Springenberg, J. T., Fiederer, L. D. J., Glasstetter, M., Eggensperger, K., Tangermann, M., Hutter, F., Burgard, W. & Ball, T. (2017). Deep learning with convolutional neural networks for EEG decoding and visualization. Human brain mapping.

@article {HBM:HBM23730,
author = {Schirrmeister, Robin Tibor and Springenberg, Jost Tobias and Fiederer,
  Lukas Dominique Josef and Glasstetter, Martin and Eggensperger, Katharina and Tangermann, Michael and
  Hutter, Frank and Burgard, Wolfram and Ball, Tonio},
title = {Deep learning with convolutional neural networks for EEG decoding and visualization},
journal = {Human Brain Mapping},
issn = {1097-0193},
url = {http://dx.doi.org/10.1002/hbm.23730},
doi = {10.1002/hbm.23730},
month = {aug},
year = {2017},
keywords = {electroencephalography, EEG analysis, machine learning, end-to-end learning, brain–machine interface,
  brain–computer interface, model interpretability, brain mapping},
}

Installation

Basics

If you don't have pip and/or git installed

sudo apt-get install python-pip git

Clone the repository

git clone https://github.com/robintibor/braindecode.git

Make the requirements

cd braindecode
make requirements

Ignore this error:

/sbin/ldconfig.real: Can't create temporary cache file /etc/ld.so.cache~: Permission denied
make: *** [scikits-samplerate] Error 1

Install Python packages

The following can be done in or outside a virtualenv. Make sure to have it activated if you want to use a virtualenv. The following installation steps can take quite long, even above an hour.

Option 1 (with Makefile):

make install

or if you want to install with user flag for pip:

make install PIP_FLAG=--user

Option 2 (with requirements.txt):

pip install -r requirements.txt
make scikits-samplerate-pip
python setup.py develop (optionally with --user)

Cudnn

Cudnn is not a strict requirement, however everything will be slower without it. Follow the instructions to install it correctly for theano: http://deeplearning.net/software/theano/library/sandbox/cuda/dnn.html

In ipython or jupyter notebook, you can check if theano is using cudnn with:

>>> import theano.sandbox.cuda.dnn; theano.sandbox.cuda.dnn.dnn_available()

If it shows True, cudnn is being used, otherwise not.

PyCuda

Atleast one person told me you need to install PyCuda to use theano on GPU :) Follow installation instructions here: http://wiki.tiker.net/PyCuda/Installation

Test installation

Start jupyter notebook in terminal and navigate to braindecode/notebooks/tutorials/Artificial_Example.ipynb. If it works, everything is fine :)

Work with real data

Create folder <repositoryfolder>/data/BBCI-without-last-runs/

Put the following files in there:

AnWeMoSc1S001R01_ds10_1-12.BBCI.mat
BhNoMoSc1S001R01_ds10_1-12.BBCI.mat
FaMaMoSc1S001R01_ds10_1-14.BBCI.mat
FrThMoSc1S001R01_ds10_1-11.BBCI.mat
GuJoMoSc01S001R01_ds10_1-11.BBCI.mat
JoBeMoSc01S001R01_ds10_1-11.BBCI.mat
KaUsMoSc1S001R01_ds10_1-11.BBCI.mat
LaKaMoSc1S001R01_ds10_1-9.BBCI.mat
MaGlMoSc2S001R01_ds10_1-12.BBCI.mat
MaJaMoSc1S001R01_ds10_1-11.BBCI.mat
MaVoMoSc1S001R01_ds10_1-11.BBCI.mat
NaMaMoSc1S001R01_ds10_1-11.BBCI.mat
OlIlMoSc01S001R01_ds10_1-11.BBCI.mat
PiWiMoSc1S001R01_ds10_1-11.BBCI.mat
RoBeMoSc03S001R01_ds10_1-9.BBCI.mat
RoScMoSc1S001R01_ds10_1-11.BBCI.mat
StHeMoSc01S001R01_ds10_1-10.BBCI.mat
SvMuMoSc1S001R01_ds10_1-12.BBCI.mat

Now you should be able to also run braindecode/notebooks/tutorials/Lasagne.ipynb.