README

This is the code for EEGWaveNet_PrototypeTraining.

Run the code

To run the code, you need to do the following:

0. Install the the requirements.

1. Download the dataset into ./Data/Das2016 (Here we use Das-2016 as an example, you can process the other datasets in the same way).

2. Read files `./Data/get_data_for_Das-2016.txt and preprocesse the data accordingly.

3. Change the "base_path" and "DEVICE" in ./utils/cfg.py according to your environment. "DEVICE" must be 'cpu' or 'cuda'.

4. Run

   • Run ./run/run0.sh to get the results of CNN.
   • Run ./run/run1.sh to get the results of DenseNet_3D. Note that "topo" must be True for this model.
   • Run ./run/run2.sh to get the results of EEGWaveNet. You can change the parameter $K$ in the paper by changing "prototype".

   You can change other parameters according to your needs.

5. Results

   • You can find the results in ./results.