This shows you the differences between two versions of the page.
— |
nlp-private:language-id-features [2015/04/22 15:15] (current) ryancha created |
||
---|---|---|---|
Line 1: | Line 1: | ||
+ | == File Level Features == | ||
+ | * minF0 | ||
+ | * maxF0 | ||
+ | * rangeF0 | ||
+ | * avgF0 | ||
+ | * duration | ||
+ | == Slice Features == | ||
+ | * seglola -- broad-class phone | ||
+ | |||
+ | === Formants === | ||
+ | * begF0, begF1, begF2, begF3, begF4, begF5 | ||
+ | * medF0, medF1, medF2, medF3, medF4, medF5 | ||
+ | * endF0, endF1, endF2, endF3, endF4, endF5 | ||
+ | * minF0 | ||
+ | * maxF0 | ||
+ | * tminF0 | ||
+ | * tmaxF0 | ||
+ | |||
+ | === Pitch === | ||
+ | * avgPitch | ||
+ | * minPitch | ||
+ | * maxPitch | ||
+ | * begPitch | ||
+ | * endPitch | ||
+ | * centerPitch -- pitch value at center of regression line | ||
+ | * pitchSlope -- from linear regression | ||
+ | * pitchSlopeStdErr -- from linear regression | ||
+ | * pitchDensity -- sample in the segment that have a pitch defined, divided by total number of samples in the segment | ||
+ | * pitchCount -- number of samples in the segment that have a pitch defined | ||
+ | |||
+ | TODO: add quadratic regression features here. | ||
+ | |||
+ | [[Category:Spoken Language ID]] | ||