Intuitive Visual Imagery Decoding for Drone Swarm Formation Control from EEG Signals

Many methods, such as a joystic and a head mounted display, are used to control unmanned aerial vehicles like drones, but in recent years, many studies are being conducted on electroencephalogram (EEG)-based brain-computer interface (BCI) control methods. The non-invasive BCI does not require surgery and has an excellent temporal resolution, which is advantageous for continuously understanding the intention of the user. However, when performing stimuli-based paradigms among several paradigms, there is a disadvantage in that users must constantly concentrate on watching external stimuli displayed on the monitor. Therefore, it is necessary to use an intuitive paradigm. Nevertheless, the intuitive paradigm has rarely been used to control the formation of swarm drones until now. The EEG-based brain-computer interface using intuitive paradigms includes motor imagery, visual imagery, and imagined speech. Thus, by comparing and analyzing four kinds of drone swarm formation control using visual imagery paradigm, we try to find a more efficient method for flight control of swarm drones. As a result, we found that the average of the 4-class classification performance was 76.4% and the highest accuracy was 83% between 8 and 13 Hz. From now on, this study will be a cornerstone for domestic and overseas related research.