Sphinx based accoustic model not work

Speech Recognition Toolkit

Brought to you by: air, arthchan2003, awb, bhiksha, and 5 others

This project can now be found here.

Sphinx based accoustic model not work

Forum: Help

Creator: Sumeet

Created: 2014-01-22

Updated: 2014-01-29

Sumeet - 2014-01-22

i have compiled Pocket sphinx demo app it working fine on android with model en-us-semi & cmu07a.dic but when i change the acoustic model and language model to hindi it doesn't work
here is hindi model
https://sourceforge.net/projects/hindiasr/files/Hindiasr/HindiASR-2.0/
which is based on sphinx3

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

Pranav Jawale - 2014-01-23

what doesn't work?

Useful links
http://cmusphinx.sourceforge.net/wiki/faq
https://sourceforge.net/p/cmusphinx/discussion/search/?q=pocketsphinx+demo

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

Sumeet - 2014-01-23

Does Pocket Sphinx require different model than Sphinx because i also tries these model does not work, apps hangs and force closes.
acoustic model hmm
https://sourceforge.net/projects/cmusphinx/files/Acoustic%20and%20Language%20Models/US%20English%20Generic%20Acoustic%20Model/
language model lm
https://sourceforge.net/projects/cmusphinx/files/Acoustic%20and%20Language%20Models/US%20English%20Generic%20Language%20Model/
these model works on sphinx but not work on pocket sphinx why?
the only model work with Pocket-Sphinx is this https://sourceforge.net/p/cmusphinx/code/HEAD/tree/trunk/pocketsphinx-android-demo/assets/models/

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

Nickolay V. Shmyrev - 2014-01-23

these model works on sphinx but not work on pocket sphinx why?

You can find details in logcat output

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

Sumeet - 2014-01-26

thank Parnav & Nickolay it works now that was the issue in phenomes but the efficiency is very low i'm saying something and it recognize different,
i am creating a new acoustic model of 20 30 words but lmtool link is not working from past few days
http://www.speech.cs.cmu.edu/tools/lmtool-adv.html
http://www.speech.cs.cmu.edu/tools/lmtool.html
both are are not working, i download Legios package but stuck in process its not working

Does HTK or julius based model work in Pocket sphinx??

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

Pranav Jawale - 2014-01-27

"new acoustic model" do you mean new language model?

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

Sumeet - 2014-01-28

yes,
i often get error on step 5 ??
http://cmusphinx.sourceforge.net/wiki/tutoriallm#arpa_model_training
i have a dictionary and want Language Model
but these links are not working
http://www.speech.cs.cmu.edu/tools/lmtool.html
http://www.speech.cs.cmu.edu/tools/lmtool-adv.html

Last edit: Sumeet 2014-01-28

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

Pranav Jawale - 2014-01-29

i often get error on step 5

What error? Upload your data.

but these links are not working

Use SRILM or other tools described at http://cmusphinx.sourceforge.net/wiki/tutoriallm#using_other_language_model_toolkits

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

Log in to post a comment.