The project is aimed at creating a Real-time Left-Right Motor Imagery Classifier using CNN. Device used is a Muse 2 brain sensing headband with 4 electrode channels - TP9, AF7, AF8, TP10. Data is streamed using the 'Mind Monitor' app.
Following codes have been implemented till now:
MotorImagery_OSC_Record -
- Record and save EEG data as CSV files from a Muse 2 headband using the MInd Monitor app and python osc module.
- Events can be configured in the rec_dictionary
MotorImagery_Training -
- Configure and train a CNN model based on 'EEG-ITNet'
- Load the CSV data recordings into a Pandas dataframe and convert into MNE epochs for training
MotorImagery_OSC_Predict -
- Make real time predictions using the trained model.