integrate speech recognition with Asterisk; simple config

[johnyjj2]

Hello!

I created acoustic model (based on speech recorded myself, list of words, phonemes etc) for my language in Sphinxtrain. I'd like to connect all of these component, i.e. PBX (Asterisk) with help of Cairo/Zanzibar with speech recognition Sphinx4. I've got some difficulties with configuring this whole Cairo/Zanzibar. I couldn't find one thing but I think person with knowledge about Asterisk should be able to help me. Their support doesn't work too well, mailing list of Zanzibar project is empty, Cairo list is rather rarely active. Fortunately they've got good on-line documentation (

http://www.spokentech.org/openivr/intro.html

oraz

http://www.speechforge.org/projects/cairo/intro.html).

I'd like to know: how to include my own acoustic model? They use default models from Sphinx, I'd like to use my own model. I guess it is just a matter of simple edition in configuration file. (The second difficulty is about the place where I need to specify my algorithm, would it be source code of Zanzibar or rather of speech recognition Sphinx4)?

I explained my problem in details here:

http://old.nabble.com/using-other-acoustic-models-in-Cairo-Zanzibar-to26879547.html

on their mailing list but there is no answer.

The answer is really important to me so I would be grateful for help!

Regards!