EEGIS - Electroencephalogram Imagined Speech dataset

This Dataset contains Imagined Speech EEG signals.

The signals were recorded from 10 participants while they were imagined saying eight different Spanish words:
- 'Sí'
- 'No'
- 'Baño'
- 'Hambre'
- 'Sed'
- 'Ayuda'
- 'Dolor'
- 'Gracias'
plus a rest state.

The words translated are 'Yes', 'No', 'Bath', 'Hunger', 'Thirst', 'Help', 'Pain', 'Thank you'.

The EEG signals were recorded using the Emotiv EPOC+ device, equipped with 14 channels and operated at a sampling rate of 128 Hz.
The data consist of chunk records of 1 second each, for that reason, data has a shape of 14 channels and 128 samples (128 Hz x 1 second) for each recording.

Also, each chunk was filtered into five frequency bands: delta, theta, alpha, beta and gamma.

Therefore, the dataset has 6 folders (raw data plus five frequency folders) where each folder contains nine folders (one corresponding to each class) where each folder have its corresponding CSV files.
Class folders encoding:
- class_0 : rest state
- class_1 : Ayuda
- class_2 : Baño
- class_3 : Dolor
- class_4 : Gracias
- class_5 : Hambre
- class_6 : No
- class_7 : Sed
- class_8 : Sí