Hi, i'm tring to train acoustic model for wake up word but in arabic , actullay i'm beginner in this field so i followed Training an acoustic model for CMUSphinx i recorded data for just this word and followed the steps but when i train
context-dependent models i dont get anything nor paramter file created and this is only what i get
when i change $CFG_CD_TRAIN = 'yes'; to no and $CFG_CI_MGAU = 'no'; to yes
it creates paramter file but when i run it it actually recognize anything as the wake up word anything i say it just recognize it as the word
so i want to know what is wrong is it because not enough data or this is not the right way to train model for wake up word ??
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi, i'm tring to train acoustic model for wake up word but in arabic , actullay i'm beginner in this field so i followed Training an acoustic model for CMUSphinx i recorded data for just this word and followed the steps but when i train
context-dependent models i dont get anything nor paramter file created and this is only what i get
when i change $CFG_CD_TRAIN = 'yes'; to no and $CFG_CI_MGAU = 'no'; to yes
it creates paramter file but when i run it it actually recognize anything as the wake up word anything i say it just recognize it as the word
so i want to know what is wrong is it because not enough data or this is not the right way to train model for wake up word ??