Hi!!!
Is there a way to get, through pocketsphinx, a sort of "success rate" or a
mark that indicates how much pocketsphinx "thinks" that the word decoded is
correct? I mean, pocketsphinx gives a mark in decoding audio? I would like to
know it because, considering the mark, i can establish if the audio decoded
it's rubbish or effectively a word!
Another question: I've got some float arrays where sound is sampled. Can i
pass directly the flot array to pockesphinx api(s) for recognition?
Thank you and sorry for my worst english.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Is there a way to get, through pocketsphinx, a sort of "success rate" or a
mark that indicates how much pocketsphinx "thinks" that the word decoded is
correct? I mean, pocketsphinx gives a mark in decoding audio? I would like to
know it because, considering the mark, i can establish if the audio decoded
it's rubbish or effectively a word!
This is called "confidence score" or "posterior probability". You can access
this value through API:
I'm sorry for wrong section about this topic. So, for the array, i've got to
convert them in shorts. And then? What function will decode it? Thank you for
the ps_get_prob function!
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi!!!
Is there a way to get, through pocketsphinx, a sort of "success rate" or a
mark that indicates how much pocketsphinx "thinks" that the word decoded is
correct? I mean, pocketsphinx gives a mark in decoding audio? I would like to
know it because, considering the mark, i can establish if the audio decoded
it's rubbish or effectively a word!
Another question: I've got some float arrays where sound is sampled. Can i
pass directly the flot array to pockesphinx api(s) for recognition?
Thank you and sorry for my worst english.
This is called "confidence score" or "posterior probability". You can access
this value through API:
No, you need to convert the data to the array of shorts
Please use Help forum to ask for help. This forum is about generic discussion
about speech recognition.
I'm sorry for wrong section about this topic. So, for the array, i've got to
convert them in shorts. And then? What function will decode it? Thank you for
the ps_get_prob function!
Please read the tutorial
http://cmusphinx.sourceforge.net/wiki/tutorialpocketsphinx
ps_process_raw