Quantitative Comparisons between Time Domain Speech Fundamental Frequency Estimation Algorithms

. T W O techn iques a r e presented here t o enable q u a n t i t a t i v e comparison o f t i m e domain fundamental f requency e s t i m a t i o n a l g o r i t h m s a g a i n s t a r e fe rence , t h a t makes use o f t h e o u t p u t f rom a laryngograph. These measures a re c a r r i e d o u t on t h e p u l s a t i l e ou tpu t s produced by t h e dev ices, where each pu lse corresponds t o an epoch o f a c o u s t i c e x c i t a t i o n due t o a v o c a l f o l d c l o s u r e . The r e s u l t s g i ven he re a r e f o r a peak-p ick ing a lgor i thm. The comparison techn iques are: 1 ) Receiver ope ra t i ng c h a r a c t e r i s t i c . Th i s i s a p l o t o f t h e p r o b a b i l i t y o f success fu l d e t e c t i o n o f a voca l f o l d c l osu re , as compared t o t h e re fe rence , a g a i n s t t h e number o f f a l s e a la rms. It i s shown t h a t t h i s measure g i v e s a c l e a r i n d i c a t i o n as t o how w e l l t h e dev ice under t e s t per forms w i t h r espec t t o t h e re fe rence , as w e l l as p r o v i d i n g a q u a n t i t a t i v e method f o r dev ice parameter o p t i m i s a t i o n . 2 ) J i t t e r d i s t r i b u t i o n . Th i s i s a h is togram o f t h e d i f f e r e n c e s i n t h e t imes o f occurence o f o u t p u t pu l ses f rom t h e re fe rence and t h e corresponding t ime -a l i gned pu lses f r om the dev ice under t e s t . Th i s measure g i ves an i n d i c a t i o n o f hovd p r e c i s e l y and c o n s i s t e n t l y dev ices a r e ab le t o l o c a t e v o c a l f o l d c l o s u r e i n s t a n t s .