Skip to main content

Table 6 Parameters with percentage error below 50 %

From: Utterance copy through analysis-by-synthesis using genetic algorithm

Branch

Parameters

 

PE

Male

Female

Voicing source

P E<10 %

AV and OQ

F0, AV, and OQ

 

10 %<P E<50 %

AH and TL

AH and TL

Cascade branch

P E<10 %

F1, F2, F4, B4, and B5

F4, FNP, B4, and B5

 

10 %<P E<50 %

F3, B3, FNP, BNP, and FNZ

F1, F2, F3, B3, and BNP