op.recognize

Speech and phoneme recognition and segmentation.

Recognize is a speech to text based on Sphinx4. It translates the incoming audio signal into text and can give dates of words and phonemes in a buffer~.

Uses JSGF grammar file and ngram model languages. No special voice learning is needed and potentially multilingual.

joeyfivecents's icon

This is great. Do you know of any patches that do the opposite? Like text to speech synthesis.

rolnxyz's icon

@joeyfivecents

I don't know if this helps. but If you use mac, you could use the shell tool and execute say through osascript

osascript -e 'say "joey five cents" using "Kathy"'

Michaud's icon

any tips on installing this for windows?

Esteban's icon

aka.speech

Olivier Pasquet's icon

> any tips on installing this for windows?< Since it is a mxj external, you should simply read the "readme.txt" file.
- Move the class files to the max class folder, then move the jar file to the max lib folder.
- You normally have to run jsapi.sh from the max lib folder. This would make a jsapi.jar file. Since I do not have time to investigate on windows' Command Prompt, I simply uploaded a jsapi.jar (works for both windows and mac):
http://www.opasquet.fr/dl/jsapi.jar.zip
This would consider you agree the terms and conditions of use from SUN...

Hugobox's icon

Hi!
The download link is down, is there another?
Thanks.

Olivier Pasquet's icon

The link does not seem to be down now. You can download it from the "download" link.

best,

theslyprofessor's icon

I keep trying to download this from your site and experiencing an unexpected end of the archive.

Thanks,
Nakul

Federico Llach's icon

Hi Olivier,
Thanks for sharing your tool. Do you know if it'll work in Max 7?
Thanks,
Federico

Will Wheel's icon

When using aka.listen, how do you get the recognized words in the Print window back into the patcher window so that it can be used to activate other things in Max?

alain's icon

Hello,
I've just installed "op.recognize" on a Windows PC for Max 7.
It seems well done as the "mxj op.recognize buffer" object is white and not brown.

But when I want to load the language model in the "file" subpatcher, I've got the following error in the Console :
"class not found !java.lang.ClassNotFoundException: edu.cmu.sphinx.model.acoustic.WSJ_8gau_13dCep_16k_40mel_130Hz_6800Hz.Model
op.recognize-> error allocating Property exception component:'lexTreeLinguist' property:'acousticModel' - component 'hub44' is missing
edu.cmu.sphinx.util.props.InternalConfigurationException: component 'hub44' is missing"

Any idea how to fix this ?

Thanks,
Alain

Anna Sartzidou's icon

Hello,
I would like to ask if op.recognize can work on windows for Max 8.
Thank you in advance,
Anna

Published

Creative Commons License