Methods to identify imagined words in continuous EEG signals.
Methods to identify imagined words in continuous EEG signals.
Speaker: Carlos Tonatiuh Hernandez Del Toro
(Universidad de Guadalajara)
Date: February 27th 2026, 16:00 (GMT-6)
Auditorium Z of CUCEI. (Youtube Link)
Abstract
In this talk, we examine the potential of imagined speech—particularly imagined words—as a control signal for Brain-Computer Interfaces (BCIs). We discuss the unique challenges in classifying such covert neural activity and present established methods for extracting discriminative features from Electroencephalogram (EEG) signals prior to training supervised machine learning classifiers. Additionally, we address the practical hurdles of processing continuous EEG streams in real-time, a critical requirement for online BCI systems.