I'm trying to adapt the english model to handle audio that has noise.
I'm following the adaptation tutorial because I don't have enough audio to train a model from scratch.
I think because of the noise in the input I'm getting quite a large number of alignment errors from BW,
Is it possible to pass in alignment data to BW? Or is there anything else I should try?
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
You should try more advanced neural network toolkit like Kaldi. As for data to train, it should not be a problem, there is so much data around these days.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
You should try more advanced neural network toolkit like Kaldi. As for
data to train, it should not be a problem, there is so much data around
these days.
I'm trying to adapt the english model to handle audio that has noise.
I'm following the adaptation tutorial because I don't have enough audio to train a model from scratch.
I think because of the noise in the input I'm getting quite a large number of alignment errors from BW,
Is it possible to pass in alignment data to BW? Or is there anything else I should try?
You should try more advanced neural network toolkit like Kaldi. As for data to train, it should not be a problem, there is so much data around these days.
Now that you just posted an Android build of Kaldi, I might be able to move
from PocketSphinx!
On Thu, 5 Sep 2019 at 18:54, Nickolay V. Shmyrev nshmyrev@users.sourceforge.net wrote: