hi,
i am trying to use pocketsphinx in xamarin android. i was able to load the library and everything seems to work weel(there are no errors) but the result is always empy.
i am working with wav file with 16000 sample rate, 16 bit encoding and mono channel.
hope someone could help :(
this is my code :
It seems you are using phonetic LM en-phone.lm.dmp instead of real lm en-us.lm.bin. It does not seem like a good idea. If you want phones, you need to use -allphone instead of -lm. If you want words, you need real language model.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
and it is working :)
when i am working with the "goforward.raw" file it is perfect but when i am working with wav file that was recorded on the phone the result is not as expected.
How can i make the process more accurate ?
thanks,
royi
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
hi,
i am trying to use pocketsphinx in xamarin android. i was able to load the library and everything seems to work weel(there are no errors) but the result is always empy.
i am working with wav file with 16000 sample rate, 16 bit encoding and mono channel.
hope someone could help :(
this is my code :
this is log for the run:
From your code it seems you are trying to decode mp4 file, you need to feed in raw PCM data, not compressed sound.
thanks for your answer,
sorry for that, it is from all my expirimats
i convert the file to wave file and still the same problem
Provide an updated log please.
hi,
posted the new log below,
i also tried the file in the windows version and it return an answer, not what i said:) but something
thanks
Hi,
this is the new line:
and this is the new log:
It seems you are using phonetic LM
en-phone.lm.dmp
instead of real lmen-us.lm.bin
. It does not seem like a good idea. If you want phones, you need to use-allphone
instead of-lm
. If you want words, you need real language model.Thanks!!
i have changed the code to:
and it is working :)
when i am working with the "goforward.raw" file it is perfect but when i am working with wav file that was recorded on the phone the result is not as expected.
How can i make the process more accurate ?
thanks,
royi
Send data to google, their recognize is much more accurate.