We are delighted to announce that the BINGO EEG datasets are now publicly available!
After completing rigorous data collection and careful preprocessing, the BINGO team has released the full multi-session EEG dataset for imagined speech decoding. This release represents a significant contribution to open science in neurotechnology providing researchers, practitioners, and machine learning developers with access to one of the most comprehensive imagined speech EEG resources yet.
The dataset includes:
- Multi-session EEG recordings collected across three separate days
- Imagined production of all NATO phonetic alphabet elements
- Standardised data formatting and detailed documentation
- Baseline methods and sample code to get you started
By making the dataset openly accessible, we aim to:
- Support reproducible research and fair comparison across methods
- Enable independent validation of new algorithms
- Kickstart innovation in EEG-based brain–computer interfaces
- Foster collaboration across neuroscience, engineering, and AI communities
The datasets are hosted on our website [link] and are accompanied by clear usage guidelines and metadata that facilitate reuse.
We hope this release accelerates research and opens new possibilities for decoding imagined speech from neural signals. Whether you are developing machine learning models or exploring neurophysiological dynamics, the BINGO dataset is now at your fingertips.
Explore the data and help decode the future of communication!