SNN for MI-EEG Classification

Forked from original https://doi.org/10.1016/j.bspc.2024.107000 "Constructing Lightweight and Efficient Spiking Neural Networks for EEG-based Motor Imagery Classification" by @XiaoJianLiao

Requirements

spikingjelly 0.0.0.0.14 pytorch 1.12.1 scikit-learn 1.5.1 scipy 1.13.1

File list

Original Scripts

Pipeline.py: An example for constructing lightweight and efficient spiking neural networks for EEG-based motor imagery classification

Train_ANN.py: Training LENet under ANN structure

Tset_SNN.py : Convert trained LENet to SNN

DataSets

BCICIV_2a_gdf: The preprocessed dataset for testing the framework

Jupyter Notebooks (for Google Colab)

SNN_for_EEG_based_Motor_Imagery_Classification.ipynb: All the scripts in one file