Re: Sphinx acoustic model to htk acoustic model

Acoustic Model Discussions

Flat

Sphinx acoustic model to htk acoustic model

User: yassing
Date: 11/11/2010 7:13 pm

Views: 5982
Rating: 7

I want to transform model from Sphinx acoustic model to htk acoustic model.

Is Speech Recognition Model Coverter used to transform Sphinx acoustic model to htk acoustic model?

But I go to website, http://sourceforge.net/projects/srmc/ , I can not find Speech Recognition Model Converter files.

How can I get Speech Recognition Model Coverter source?

--- (Edited on 11/11/2010 7:13 pm [GMT-0600] by yassing) ---

Re: Sphinx acoustic model to htk acoustic model

User: kmaclean
Date: 11/12/2010 12:17 am

Views: 56
Rating: 6

>How can I get Speech Recognition Model Coverter source?

http://srmc.cvs.sourceforge.net/viewvc/srmc/usmc/

--- (Edited on 11/12/2010 1:17 am [GMT-0500] by kmaclean) ---

Re: Sphinx acoustic model to htk acoustic model

User: yassing
Date: 11/12/2010 2:04 am

Views: 208
Rating: 3

Thank you.

I get USMC, and build USMC using VC2008.

Finally, I get s3mod2htk.exe.

Do you know how this excute file use?

--- (Edited on 11/12/2010 2:04 am [GMT-0600] by yassing) ---

Re: Sphinx acoustic model to htk acoustic model

User: dara
Date: 1/13/2011 11:41 am

Views: 2677
Rating: 3

Hi there,

First of all, via command terminal go to the directory where the .exe is, then type in the following command:

s3mod2htk -moddeffn <path of your mdef> -meanfn <path of your means> -varfn <path of your variances> -tmatfn <path of your transition matrices> -mixwfn <path of your mixture weights> -binary <yes/no> -mach_endian <big/little> -htkmmffn <path of your MMF output> -htkmacrofn <path of your macros output>

Note: remove the "<>"

Or just type in "s3mod2htk" (without "), and it'll give you all the options you need.

regards,

Dara

--- (Edited on 1/13/2011 11:41 am [GMT-0600] by dara) ---

Previous • Next •


Username	Password