Click here to register.

Frequently Asked Questions

Nested
How to implement a continuous speech recognition using HTK's HDecode
User: kmaclean
Date: 1/1/2010 11:13 pm
Views: 5183
Rating: 2    Rate [
]

From the HTK mailing list archives:

> Hi everyone,
>
> I am trying to implement a continuous speech recognition in Spanish 
> Language. I followed the indications to make the tri-gram language 
> model presented in the HTK book. I am using the HDecode tool. My 
> firsts results are very poor (WER >= 35%). So I tuned various 
> parameters in HDecode, but know I need to make tuning of the language 
> model parameter. Any user can help me in that? for example, what 
> parameter in the to language model generation are recommended to 
> continuous speech?

Did you try to follow htk wsj1 reciept? It has almost everything
required I think:

http://www.inference.phy.cam.ac.uk/kv227/htk/
http://www.inference.phy.cam.ac.uk/kv227/lm_giga/

with all beams used and lm factors. Though for really large vocabulary lm factor should be smaller (around 6-8).

Re: How to implement a continuous speech recognition using HTK's HDecode
User: vaibhavi trivedi
Date: 3/13/2013 7:02 am
Views: 218
Rating: -15    Rate [
]

how to implement a continuous speech recognition using HTK's ???

Re: How to implement a continuous speech recognition using HTK's HDecode
User: vaibhavi trivedi
Date: 3/13/2013 7:38 am
Views: 249
Rating: -4    Rate [
]

how to implement a continuous speech recognition using HTK's ???

PreviousNextAdd