/FeatureShift

Primary LanguagePython

Arousal Mental State Classification

This is a research project to do binary classification of EEG arousal mental state using deep learning

Dataset

The dataset is collected with Muse EEG headband and HTC Vive pro VR headset.
  • Number of subjects: 18
  • Classes: 2 (arousal vs calm)
  • Sampling rate: 256 Hz
  • Experiment protocol: (1 min calm + 1 min arousal + 5 s waitting) * 3
  • Total duration for each subject: 6 mins

    • About the code

    Requirement:

    • pytorch == 1.2.0 or above
    • h5py == 2.9.0
    • scipy == 1.3.1
    • numpy == 1.16.4

    How to use

    Please use Main.py as the starting script

    For training, use "[object of the class Train].set_parameter()" to modify the parameters. Please only use .Leave_one_session_out() to do subject-dependent classification task

    All the result will be save in to a txt file named "result_leave_one_session_out_record.txt" at the current directory of the script. Also, the Acc of training and evaluation can be found in "Result_model/Leave_one_session_out/history/" under the same directory.

    Paper Citation

    If you find the code is useful, please cite our work:

    Author...,Decoding Emotional Arousal Mental State Using Frequency Correlation Graph Convolutional Neural Networks,Source

    Please try our best to get a high quality publication!