Hi. I created .grammar file like this below;

S : NS_B DIGIT_LOOP NS_E
DIGIT_LOOP: DIGIT_LOOP DIGIT
DIGIT_LOOP: DIGIT

So. When run julius i expect to recognize my voice up to <sil> like up to end of speech. However, it only prints one word and prints wheter my first word or immediate ones or last and it changes . It is like for continuous speech it takes randomly. My grammar has ability to continuous speech recognition from given .voca file. Okey, it is wait until i finish talking. By if i speech only 1 word, it recognize trully.

But why it doesnt print more than 1 word as output?

--- (Edited on 12/19/2015 5:48 pm [GMT-0600] by Visitor) ---