ICARECS-2026 - EEG -DRIVEN MOTOR IMAGERY CLASSIFICATION USING DEEP NEURAL NETWORK FOR NEUROREHABILITATION APPLICATIONS
i-manager's Journal on Pattern Recognition · 20.06.2026 · 13 (1) · 1
DOI: 10.26634/jpr.13.1.11292 ↗Kurz gefasst
Fachartikel in der Zeitschrift i-manager's Journal on Pattern Recognition, online erschienen am 15.06.2026. Verfasst von Sathiyanathan N.
Abstract (Originalsprache)
The recognition of motor imagery (MI) event-related potential EEG in brain–computer interface (BCI) is still a difficulty problem owing to the low signal-to-noise ratio and strong non-stationary characteristic of EEG signal which varied with time, environment and individual. In this paper, a motor imagery classification framework based on a modified Improved Feature Network (IFNet) is proposed. The raw EEG signal passed through band-pass filtering and structured trial segmentation for extracting the motor imagery epochs. The EEG epochs were put into the enhanced IFNet for extracting the characteristic features. The convolutional layers in the network can deeply mine the features of different EEG epochs in spatial–temporal domain. The batch normalization and dropout could help to alleviate the deviation of network and enhance the robustness of the network. Unlike the general approach of manually selecting the appropriate features for the BCI tasks, the proposed network can automatically learn the relevant features for BCI performance. The experimental results show that the proposed framework can achieve the better BCI performance, and faster and more stable convergence of the network, and also with a larger inter-class separability.
Zitieren
APA
N, S. (2026). ICARECS-2026 - EEG -DRIVEN MOTOR IMAGERY CLASSIFICATION USING DEEP NEURAL NETWORK FOR NEUROREHABILITATION APPLICATIONS. i-manager's Journal on Pattern Recognition, 13(1), 1. https://doi.org/10.26634/jpr.13.1.11292
Vancouver
N S. ICARECS-2026 - EEG -DRIVEN MOTOR IMAGERY CLASSIFICATION USING DEEP NEURAL NETWORK FOR NEUROREHABILITATION APPLICATIONS. i-manager's Journal on Pattern Recognition. 2026;13(1):1. doi:10.26634/jpr.13.1.11292
BibTeX
@article{N2026,
author = {N, Sathiyanathan},
title = {ICARECS-2026 - EEG -DRIVEN MOTOR IMAGERY CLASSIFICATION USING DEEP NEURAL NETWORK FOR NEUROREHABILITATION APPLICATIONS},
journal = {i-manager's Journal on Pattern Recognition},
year = {2026},
volume = {13},
number = {1},
pages = {1},
doi = {10.26634/jpr.13.1.11292}
}
EUVIMED