| Summary: | 碩士 === 南台科技大學 === 電機工程系 === 102 === Using cepstrum analysis and Bayesian inference of SSVEP-based BCI is proposed in this study. Power cepstrum analysis is adopted to obtained evoked potential parameters. For different reaction evoked potentials established Gaussian mixture module. Finally, through Bayesian inference completion of EEG response signal classification.
Experimental results show that, the performance of the method proposed in this study is better than the previous BCI research. At different excitation frequencies, the proposed method of recognition rate is 94.09% better than 91.73% CCA method. The method of this study will not because the training data different affect system recognition rate. The method of this study were less likely to be affected by noise. And reached an user independent of the characteristics.
Keywords: Brain computer interface, electroencephalogram, Steady-state visual evoked potential, Cepstrum analysis, Bayesian inference
|
|---|
