Affordable brain-computer interface (BCI)
Supervisor : Mojtaba Ahmadi
Team size: Minimum , Maximum
| CSE | SE | Comm | Biomed | EE | Aero | Special |
|---|---|---|---|---|---|---|
| Yes | No | No | No | No | No | No |
Description
Our goal is to leverage consumer-grade EEG devices such as the Muse 2, the system will use a P300 speller interface and text input through neural responses to particular visual stimuli (e.g., letters). Individual training is required as brainwave patterns are unique across users. Prototyping cursor controls via motor imagery is also possible as well as integrating machine learning (though may be beyond scope). This project aims to provide an alternative to conventional peripherals, while prioritizing accessibility and cost.
While the initial project focuses on a "mind-to-text" speller, the system's architecture is highly extensible and can be expanded to fully support desktop navigation. Going even further, we can integrate EEG-controlled navigation of a Raspberry Pi-powered car for example. Technical challenges would include signal latency (especially with consumer grade EEG's) and user training, which would be analyzed in consideration with noise.