Hi, I'm trying to use a FSG with a list of contact names in it. However, I
keep getting the following messages:
INFO: fsg_model.c(325): Adding alternate word transitions (natasha,NATASHA(2))
to FSG
INFO: fsg_model.c(358): Added 1 alternate word transitions
The problem, is that I'm trying to flite to say the contacts name back and
keep ending up with it saying "natasha(2)". Even though the literal phrase
"natasha(2)" isn't in my grammar file.
Any insight would be greatly appreciated.
Thanks! Andrew
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
(2) here means alternate pronunciation variant from your dictionary. If you
don't need alternative variants, you can remove numbers in braces in
recognized string.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
There is no way to force dictionary without alernatives and no need to do
that. Alternative pronuciations improve recognition rate. You should better
remove them in recognizer output as I wrote above.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Anonymous
-
2010-07-08
Can you remove the (2) from the dictionary?
For example:
THE DH AH
THE DH IY
instead of
THE DH AH
THE(2) DH IY
I am trying to create a simple command and control application. Or is the best
way to do this have both THE and THE(2) as terminal symbols? There is no need
for me to know of the difference in the two pronunciations, only that the word
THE was spoken.
Thanks!
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
There is no need for me to know of the difference in the two pronunciations,
only that the word THE was spoken.
Result returned to you by ps_get_hyp call contains no information about
pronuciation variants, it's just a string of recognized words. Sorry, I don't
really get your problem here. Probably you need to explain more.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Anonymous
-
2010-07-13
So, if THE is pronounced DH IY then hyp will be THE? or THE(2)? I need to
parse the string of recognized words and will need to distinguish between,
perhaps, THE(2) and THERE.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi, I'm trying to use a FSG with a list of contact names in it. However, I
keep getting the following messages:
INFO: fsg_model.c(325): Adding alternate word transitions (natasha,NATASHA(2))
to FSG
INFO: fsg_model.c(358): Added 1 alternate word transitions
The problem, is that I'm trying to flite to say the contacts name back and
keep ending up with it saying "natasha(2)". Even though the literal phrase
"natasha(2)" isn't in my grammar file.
Any insight would be greatly appreciated.
Thanks! Andrew
(2) here means alternate pronunciation variant from your dictionary. If you
don't need alternative variants, you can remove numbers in braces in
recognized string.
Thanks for the reply. Do you know if there is a way to surpress these (i.e.,
(2)) in lmtools?
There is no way to force dictionary without alernatives and no need to do
that. Alternative pronuciations improve recognition rate. You should better
remove them in recognizer output as I wrote above.
Can you remove the (2) from the dictionary?
For example:
THE DH AH
THE DH IY
instead of
THE DH AH
THE(2) DH IY
I am trying to create a simple command and control application. Or is the best
way to do this have both THE and THE(2) as terminal symbols? There is no need
for me to know of the difference in the two pronunciations, only that the word
THE was spoken.
Thanks!
Result returned to you by ps_get_hyp call contains no information about
pronuciation variants, it's just a string of recognized words. Sorry, I don't
really get your problem here. Probably you need to explain more.
So, if THE is pronounced DH IY then hyp will be THE? or THE(2)? I need to
parse the string of recognized words and will need to distinguish between,
perhaps, THE(2) and THERE.
the hypothesis will be just "THE" without (2). I don't really understand your
last statement about THERE