The frequency recognition performance of short-time SSVEPis studied here. A quantitative analysis CCA based targetidentification methods for short-time SSVEP-based BCIs isillustrated. The aim is to improve the classification accuracy ofshort-time SSVEP for BCI implementation. The length of prefiltered(6 – 120Hz) short-time signal is artificially increased tofit for applying CCA. The proposed method produces betterrecognition accuracy than standard CCA and MsetCCA. Fig. 3shows that the proposed method exhibits higher ITR thanstandard CCA and MsetCCA. To improve the recognitionaccuracy of the proposed method, the short-time data ofmultichannel SSVEP is reformed to increase its length. Theunderlying assumption of making such modification is that theCCA method works better for higher length of data, whereas,the selected short-time SSVEP is not adequate in term of lengthfor standard CCA. The comparisons in terms of recognitionand ITR among standard CCA, MsetCCA and proposedmethod of short-time SSVEP of all subjects obtained withdifferent time window from 0.25s to 1.0s are illustrated. In bothcomparisons, proposed method achieves higher performancethan standard CCA and MsetCCA method for all TWs.A novel method for frequency recognition of short-timeSSVEP signals is introduced here. It is observed that theperformance of the proposed method is better than that of thestandard CCA and MsetCCA for a wide range of the length ofshort-time SSVEP signals. The underlying contention ofimproved performance of proposed method is that the artificialincrease (by concatenation) of short-time SSVEP dataset worksbetter for frequency recognition than directly using short-timeSSVEP dataset using CCA based approach. The experimentalresults with EEG data from 5 healthy subjects demonstratedthat the proposed method outperforms the standard CCA andMsetCCA method of recognition accuracy also informationtransfer rate (ITR) for short-time SSVEP based BCIsimplementation. At the same time, the method is more suitableto implement BCI with lower calibration time.

x

Hi!
I'm Isaac!

Would you like to get a custom essay? How about receiving a customized one?

Check it out