Research on High-Instruction-Set Visual Brain-Computer Interface Based on Eye-Tracking Monitoring Submission

In recent years, visual-based Brain-Computer Interface (BCI) systems have gained significant attention due to their high Information Transfer Rate (ITR). In practical applications, there is a growing demand for large instruction set BCI systems to support more complex commands. However, users may experience fatigue-related issues during prolonged engagement in visual tasks, which negatively impacts the modeling accuracy of BCI systems. To address the issue of signal degradation caused by user subjective intentions that is difficult to detect and process, we propose a method which can monitor user attention and optimizes signal quality in offline data processing. Under the monitor of eye tracker, this method employs Exponentially Weighted Moving Average (EWMA) and Simple Moving Average (SMA) to calculate fixation points, aligns them with the target range, and filters the electroencephalogram (EEG) signals of the current trials. Offline results demonstrate that the average accuracy, after optimizing with EWMA and SMA, is 87.31% and 86.86% respectively, while the average accuracy is 80.81% for the raw signals. This paper demonstrates that monitoring user subjective intention decay can improve the accuracy of offline models and provides a new method for taking user performance into consideration in the development of BCI applications in the future.