HFD-CCA for SSVEP classification in Mobile Environment
Himadri Panthadas, Mohammed Imamul Hassan Bhuiyan · 2024
In this paper a high frequency domain filter in conjunction with canonical correlation analysis (HFD-CCA) method is proposed for the classification of Steady state visually evoked potential (SSVEP) in mobile conditions. A publicly available dataset comprising of mobile EEG data of 23 individuals at four different speeds (0, 0.8, 1.6 and 2 m/s) is used in the study. It is demonstrated that with the proposed approach significant improvement in accuracy (2.5%) can be achieved in a highly mobile environment as compared to using CCA alone. It is further shown that across different speeds, similar or better accuracy (0.58 to 2.74%) can be obtained as compared to recently introduced multivariate iterative filtering-based approach. This indicates the potential of the presented method in brain computer interfacing within a mobile context.