Skip to main content
. Author manuscript; available in PMC: 2013 Sep 13.
Published in final edited form as: J Neural Eng. 2011 Jul 13;8(4):046028. doi: 10.1088/1741-2560/8/4/046028

Table 6.

Confusion matrix of decoding vowels for overt speech. The column labels correspond to the predicted vowels pair in a given trial. The row labels correspond to the correct vowels pair. The values in each cell give frequencies in percent and the standard deviation calculated across subjects. The best vowel pair /u:/ is marked in red.

Accuracy (%) /ε/ /æ/ /i:/ /u:/
/ε/ 27.72±7.92 18.76±6.27 24.43±8.29 29.09±4.55
/æ/ 21.38±11.83 34.09±16.00 23.79±9.45 20.74±5.78
/i:/ 19.82±6.92 17.88±8.22 38.80±7.71 23.48±8.31
/u:/ 19.74±6.38 19.64±4.77 21.70±7.36 38.91±7.68