Mental Workload Detection Based on Electroencephalograms

Summary

We provide some code to create the input data for training Mental Workload detectors based en Electroencephalogram (EEG) signals.

Requirements

• Dataset: From the "Dataset to predict mental workload based on physiological data", download the EEG data from the N-back test or the Heat-The-Chair game.

• Python 3.8+, Pandas 1.4+, Pyarrow 8.0+, and other minors.

Preparation

• In the root of your project, make sure to create two folders: /data and /input_features.

   ~/    
      /data/     
      /input_features/

• Put the donwloaded dataset in /data.
• At the end, the generated input data for your model will be located in /input_features

Running the code

1. Select different options of data preparation by editing directly the file.

   eeg_config.py   

The options you can change are: the dataset, the window_size, the window_overlapping, and the filtering method.

2. Next, run the script

   python main_create_input_data.py

Further information

You can visit the web of our Research Group.