/dn3

The fusion of MNE and PyTorch for accelerated deep-neural-network based BCI-systems and Neurophysiology signal analysis.

Primary LanguagePythonBSD 3-Clause "New" or "Revised" LicenseBSD-3-Clause

DN3 - Deep Neural Networks for Neuro-physiology

Bridge for training deep neural-network models with neuroscientific data managed using MNE.

Focused on:

  • Minimizing boilerplate for DNN powered BCI classifiers and processors
  • Rapid integration and extension to new datasets by providing a yaml interface to dataset construction
  • Platform for accessing state-of-the-art
    • Architectures (potentially with pretrained weights)
    • Pre-processing and data transformations

See guides and documentation at: https://dn3.readthedocs.io/en/latest/

Associated pre-print (article under review) can be found at: https://www.biorxiv.org/content/10.1101/2020.12.17.423197v1

Please consider citing the above in any scholarly work that uses this library.

Requirements:

  • python >= 3.5
  • pytorch >= 1.3
  • mne >= 0.20
  • pyyaml
  • pyyaml-include
  • numpy
  • pandas
  • tqdm