Comparative Analysis of VGG19, ResNet50, and GoogLeNet Inception Models for BCI

A brain-computer interface (BCI) is a system that allows an individual to communicate with computer or other external device using brain activity alone. Electroencephalography (EEG) common approach for detecting in BCI systems, as it provides non-invasive measure of high temporal resolution. Deep learning methods, such convolutional neural networks (CNNs), have been applied the analysis EEG data applications, and shown promising results terms accuracy reliability. In this literature survey, we reviewed use CNNs classification real time identify specific commands. Challenges development effective deep methods include variability signals across individuals dimensionality data. Further research needed address these challenges improve reliability systems methods.