Design and Implement the Continuous Flickering SSVEP-BCI in Augmented Reality

Abstract

Brain-computer interface (BCI) provides a new way to express our minds without peripheral nerves and muscles. In this work, a process control recognition method based on continuous flickering is proposed to output continuous commands. Phase matching method was applied in this work, matching the test data with the template with the same phase to improve the accuracy. We also put forward a high tolerance criterion and give a new definition to the recognition result of attention shift period. We first conducted a screen-based continuous flickering feasibility verification experiment using correlation component analysis algorithm (TRCA) method, and the recognition accuracy reached 85% and 90% under high tolerance criterion. The average offline simulation information translates rate (ITR) was 455.8 bit/min, and the highest ITR reached 524.7bit/min, which certificated the feasibility. Furthermore, we carried out a drone control experiment based on augmented reality (AR)-BCI using extended filter bank canonical correlation analysis (extended-FBCCA), and achieved an average accuracy of 90% and ITR of 49.1 bit/min, having better performance than repetitive visual stimulus (RVS) with intervals.