Innovative combination of covariance analysis-based sliding time window and task-related component analysis for steady-state visual evoked potential recognition
Jiaofen Nan,Siyuan Zhang,Duan Li,Kaifan Zhang,Chuang Han,Yinghui Meng,Tanxin Zhu
DOI: https://doi.org/10.1007/s10586-024-04492-6
2024-05-08
Cluster Computing
Abstract:Because of significant individual differences of brain signals, ensuring accuracy and information transfer rate (ITR) remains a challenge for SSVEP recognition. To address this issue, this study combines innovatively sliding time window method based on covariance analysis with the Task-related Component Analysis (TRCA). Moreover, the adaptive analysis technique is used to match different individuals by adjusting the initial length of time window and changing correlation analysis in TRCA. The results reveal that the smaller the initial time window length, the better the SSVEP recognition performance. With the fixed time window, TRCA using canonical correlation analysis (CCA) exhibits lower accuracy and smaller ITR compared to TRCA using Pearson correlation coefficient (PCC). However, under the sliding time window, TRCA with CCA finishes identification with 97% accuracy comparable to TRCA with PCC, and gets ITR up 20.7%. Furthermore, its accuracy and ITR are better than CCA, FBCCA, ABFCCA, and FBCCA-DW. The above results indicate that, during SSVEP identification, the sliding time window method based on covariance analysis is better paired with CCA than PCC, and the fixed time window method is better paired with PCC than CCA. Code is available.
computer science, information systems, theory & methods