Research Paper
Dynamic wavelet-based augmentation for enhanced EEG-based imagined speech classification
Computers in Biology and Medicine6/18/2026·06/18/26🌐 Asia
Summary
This paper introduces a dynamic wavelet-based data augmentation method for imagined speech decoding using EEG. The approach addresses limited training data challenges in EEG-based speech BCI by generating synthetic samples that preserve spectral-temporal characteristics of neural signals.
Why it matters
Imagined speech BCIs hold promise for restoring communication in locked-in patients, but limited training data has impeded progress. This augmentation strategy could accelerate development of practical speech BCIs by improving classification accuracy with smaller datasets.
#EEG#BCI#AI#Neural Interfaces
Want the full NeurotechMag Pro intelligence database?
Access 800+ papers, trials, patents, funding rounds and news items — updated daily.
Start free trialAlready a member? Sign in