IIST BCI Dataset-11: Male and Female EEG Dataset for Malayalam Vowels and Consonants

Brain-Computer Interface (BCI) technology enablesdirect communication between the brain and external devices byinterpreting neural signals. However, most BCI research focuseson widely spoken languages, and there is a lack of EEG datasetsfor regional languages such as Malayalam, which is spoken by40 million people in Kerala, an Indian state. Such datasetsare crucial for developing BCI solutions for neurodegenerativepatients who communicate in their native language.To address the research gap, we created, an EEG dataset forMalayalam language speech processing. The dataset includes15 vowels and 36 consonants, recorded from 3 male and 3female native Malayalam speakers, each having different regionalaccents. EEG signals were recorded using the OpenBCI Cytonboard, an 8-channel EEG acquisition system, with electrodesplaced following the international 10-20 system to ensure optimalsignal capture. Each letter was recorded with 10 trials in bothvocal (speaking) and subvocal (thinking) conditions, making thedataset useful for speech and subvocal speech recognition tasks.This dataset can support Machine Learning (ML) and DeepLearning (DL) research in speech recognition, brain-computerinterfaces, and neurorehabilitation applications. It contributesto indian language research and can help improve communi-cation for Malayalam-speaking patients with neurodegenerativeconditions.