Im very new to this library and willing to use it for voice recognition. I
have downloaded and build the pocketsphinx_continuous.exe and other dlls. But
after that how do I test voice recognition. I searched a lot but not getting
any proper tuts. Please guide me.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
We can use Pocketsphinx for voice recognition right ?
Yes, you can
"This manual is intended for developers who need to apply speech technology
in their applications, not for speech recognition researches.". Then what this
line trying to say us.
There are people whos speciality is to do speech recogniton research, imrove
recognition algorithms and write scientific papers about that. They will find
this manual boring. So the line says that scientists shouldn't read the
tutorial
There are developers who implement the applications to sell and share them.
They are the target audience of the tutorial.
I presume you are not scientist so you can proceed without any hestitation
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
i have same problem with you . i asked for this. i think models in the
pocketsphinx-0.6.1\model folder. i tried but i couldnt add them to
pocketsphinx.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
You need to provide some bare minimum information to pocketsphinx_continuous
in order to be able to run it. The least you need to specify are following:
A suitable model.
pocketsphinx comes with a very good model for americal english and that is
available in model\hmm\en_US folder and is called hub4wsj_sc_8k.
A dictionary file.
pocketsphinx comes with an english dictionary by the name of cmu07a and it is
available in model\lm\en_US file name is
cmu07a.dic
A grammer file or a language model.
For a starter you can create a language with an online cmu tool (Check some
tutorial on how to do it). Once you have these, following should be the
command line to run it.
pocketsphinx_continuous -hmm ..path to \hub4wsj_sc_8k -dict ..path to
cmu07a.dic -lm ..path of the language model file
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
hypheni how can you run from command prompt. i have never run from there could
you tell steps and what you write to command promt? for example my hmm file
path is
C:\Users\user\Documents\Visual Studio
2008\Projects\pocketsphinx-0.6.1\model\hmm\en_US\hub4wsj_sc_8k
what will be my command?
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
yes you can use this grammar file. Remember that this is not a "language
model" file so you'll have to change your
third argument from -lm to -jsgf as follows:
pocketsphinx_continuous -hmm ..path to \hub4wsj_sc_8k -dict ..path to
cmu07a.dic -jsgf ..path of the grammar file
Now you can save your file as .jsgf file and if you speak coffee, tea or milk,
it will recognize.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Now Im getting Runtime error ie: Windows error reporting window, saying
"pocketsphinx_continuous encountered a prob and needs to be closed........."
what could be the reason ?.
And @ ayhankm
cmd line argument through dos means you have to locate your directory in which
pocketsphinx_continuous resides by cd command. then use as follows..
Im very new to this library and willing to use it for voice recognition. I
have downloaded and build the pocketsphinx_continuous.exe and other dlls. But
after that how do I test voice recognition. I searched a lot but not getting
any proper tuts. Please guide me.
Please read the tutorial http://cmusphinx.sourceforge.net/wiki/tutorial
Yes I did. Whats the meaning of this line.
"This manual is intended for developers who need to apply speech technology in
their applications, not for speech recognition researches."
We can use Pocketsphinx for voice recognition right ?. Then what this line
trying to say us.
Yes, you can
There are people whos speciality is to do speech recogniton research, imrove
recognition algorithms and write scientific papers about that. They will find
this manual boring. So the line says that scientists shouldn't read the
tutorial
There are developers who implement the applications to sell and share them.
They are the target audience of the tutorial.
I presume you are not scientist so you can proceed without any hestitation
Yes Im not a scientist. LOL
Anyways I get the following info & error while running pocketsphinx_continuous
from command line.
Yes, INFO notifies you what is happening and ERROR gives you information where
the problem is and what to do.
Yes. I have not added any models. What are these models for and how to add
them to pocketsphinx
Hi hypheni. Did you solve your problem?
No. Please guide me how to solve it.
i have same problem with you . i asked for this. i think models in the
pocketsphinx-0.6.1\model folder. i tried but i couldnt add them to
pocketsphinx.
Thats okay. The model folder. But there are some models available for
download. You also getting same info and error as of mine ?
yes im running from visual studio 2008 and im getting same info and error as
yours. everthing is same with the fifth message in this page.
Okay. Im using VS2005 here. I think we need to pass some arguments, without
arguments all these are coming.
yes maybe. we should think path problem too. im looking now there maybe we can
find answer there : http://cmusphinx.sourceforge.net/wiki/tuturialpocketsphin
x .
Im on this link from 2days. No solutions till now.
last hope was this link. we should wait until nickolay answer this. Then we
can discus this again. if i found something i will share from here.
Hi Guys,
You need to provide some bare minimum information to pocketsphinx_continuous
in order to be able to run it. The least you need to specify are following:
pocketsphinx comes with a very good model for americal english and that is
available in model\hmm\en_US folder and is called hub4wsj_sc_8k.
pocketsphinx comes with an english dictionary by the name of cmu07a and it is
available in model\lm\en_US file name is
cmu07a.dic
For a starter you can create a language with an online cmu tool (Check some
tutorial on how to do it). Once you have these, following should be the
command line to run it.
pocketsphinx_continuous -hmm ..path to \hub4wsj_sc_8k -dict ..path to
cmu07a.dic -lm ..path of the language model file
Yes that what we need. Let me test all these settings.
Is it a sample grammar file ?..
should I save this as .jsgf file as pass as 3rd argument. And if I say coffee
or tea or milk it will recognize it ?
After all these Im getting the followings:
hypheni how can you run from command prompt. i have never run from there could
you tell steps and what you write to command promt? for example my hmm file
path is
C:\Users\user\Documents\Visual Studio
2008\Projects\pocketsphinx-0.6.1\model\hmm\en_US\hub4wsj_sc_8k
what will be my command?
Is it a sample grammar file ?..
JSGF V1.0;
grammar drinks;
public <drinks> = coffee | tea | milk
pocketsphinx_continuous -hmm ..path to \hub4wsj_sc_8k -dict ..path to
cmu07a.dic -jsgf ..path of the grammar file
Now you can save your file as .jsgf file and if you speak coffee, tea or milk,
it will recognize.
is anyone can explain how i can run pocketsphinx_continuous from command
prompt? it is a urgent sitiuation for me.
Thanks.
this assignement was quite helpful to me..
http://www.speech.cs.cmu.edu/15-492/assignments/hw1/index.html
Now Im getting Runtime error ie: Windows error reporting window, saying
"pocketsphinx_continuous encountered a prob and needs to be closed........."
what could be the reason ?.
And @ ayhankm
cmd line argument through dos means you have to locate your directory in which
pocketsphinx_continuous resides by cd command. then use as follows..
all these should be in same line. path should be replaced by your ones.