EEG-based AI-BCI Wheelchair Advancement: Transformer-Based Learning with Motor Imagery for Brain Computer Interface

arXiv cs.AI · 6 hours ago

Researchers developed TFormerEEG, a Transformer-based deep learning model for classifying motor imagery from EEG signals to control a wheelchair via brain-computer interface. The model achieved 93.04% test accuracy and 91.18% mean cross-validation accuracy when trained on 19x200 segmented EEG arrays sampled at 200Hz, outperforming baseline models like XGBoost and EEGNet. The system enables wheelchair navigation through detected right and left-hand motor imagery, with a Tkinter interface for simulation.