The Effect of Harmonics Count on SSVEP-Based BCI Results


Kancaoglu M., KUNTALP M.

Innovations in Intelligent Systems and Applications Conference (ASYU), İzmir, Turkey, 31 October - 02 November 2019, pp.110-113 identifier identifier

  • Publication Type: Conference Paper / Full Text
  • Volume:
  • Doi Number: 10.1109/asyu48272.2019.8946362
  • City: İzmir
  • Country: Turkey
  • Page Numbers: pp.110-113
  • Keywords: eeg, bci, ssvep, cca, lasso
  • Dokuz Eylül University Affiliated: Yes

Abstract

SSVEP is a BCI method based on visual stimuli selection. Compared to the well-known PSDA method, CCA and LASSO are feature reduction methods commonly used in SSVEP based Brain-Computer Interface systems due to their less noise impact. These methods require the creation of vectors with their harmonic frequencies. The number of harmonics used for SSVEP affects the results. In this study, the accuracy with 2,3 and 4 harmonics in real-time was investigated using public SSVEP data and different filters using all three SSVEP methods.