Silent speech BCI: Test of DIVA hypothesis

Ueno S., Yamazaki T., Kamijo K., YAMANOİ T.

5th International Symposium on Computational Intelligence and Industrial Applications, ISCIIA 2012, Sapporo, Japan, 20 - 23 August 2012, (Full Text) identifier

  • Nəşrin Növü: Conference Paper / Full Text
  • Çap olunduğu şəhər: Sapporo
  • Ölkə: Japan
  • Açar sözlər: BCI, EEG, Formant Frequency, Kalman Filter, Silent Speech
  • Açıq Arxiv Kolleksiyası: Konfrans Materialı
  • Adres: Bəli

Qısa məlumat

We have developed a new single-Trial-EEG-based BCI using silent speech tasks. This BCI consists of two phases. In the first phase, according to DIVA hypothesis, the relationship between EEGs, recorded during actual speech, whose dipoles were located at the premotor cortex, and formant frequencies of the speech signals is modeled by a Kalman filter. Then, in the second phase, the Kalman filter with EEGs recorded during silent speech predicts "spectrograms" of the silent speech. This study will test the above hypothesis.