Brain computer interface for object selection [Nesne seçme için beyin bilgisayar arayüzü]

Official URL: https://dx.doi.org/10.1109/SIU61531.2024.10600956

Brain-computer interfaces (BCIs) enable computer command generation through brain signals, and are applicable in diverse fields like robot control and neuromarketing. As they offer high signal-to-noise ratios, Electroencephalogram (EEG) based BCIs utilizing steady-state visual evoked potentials (SSVEP) are particularly effective. BCIs find important applications in defense technologies, with significant research investments globally. Our study presents an EEG SSVEP BCI system for object selection in defense Unmanned Aerial Vehicles (UAV) videos. By this means, we aim to facilitate the UAV control, improve the reaction time of the operator, and enable simultaneous execution of more number of functionalities. Differing from existing literature, our study focused on UAV videos (in addition to civil multimedia videos); and conducted in this context a comparison of several prominent SSVEP recognition algorithms from the literature that have been previously developed for different systems. In our performance evaluations, approximately up-to 80% top-2 selection accuracy was observed. Lastly, for a performance enhancement, we propose a new dynamic frequency tagging method.