I have increased the recognition rate of Pocketsphinx through adaption of the
acoustic model I trained by 10 %, but the rate still don't meet my use case.
Now I want to use more data for the adaption.
So my question is what do I have to consider for the choice of texts for the
adaption? In other words: How do you evaluate the suitability of a text for
adaption of an acoustic model?
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
The actual set of sentences you use is somewhat arbitrary, but ideally it
should have good coverage of the most frequently used words or phonemes in the
set of sentences or the type of text you want to recognize.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I have increased the recognition rate of Pocketsphinx through adaption of the
acoustic model I trained by 10 %, but the rate still don't meet my use case.
Now I want to use more data for the adaption.
So my question is what do I have to consider for the choice of texts for the
adaption? In other words: How do you evaluate the suitability of a text for
adaption of an acoustic model?
Hello. Thanks for your question, please avoid posting to multiple forums at
once. Choose most appropriate one.
Tutorial has the following advise on adaptation text:
http://cmusphinx.sourceforge.net/wiki/tutorialadapt
The actual set of sentences you use is somewhat arbitrary, but ideally it
should have good coverage of the most frequently used words or phonemes in the
set of sentences or the type of text you want to recognize.