Hi,
I'm very new to speech recognition and pocket sphinx; but having read the tutorials, I'm still not sure how to approach my problem.
My goal is to use a throat microphone with pocketsphinx, but the issue is that the signal outputted by the throat microphone is different from a traditional microphone. In other words, recognition is not very accurate.
I don't know if I need to create a separate language model (even though I'll still be using the same english language) or if I need to try train an acoustic model or just try to adapt the existing english model.
Any insight or intuition would be greatly appreciated. I just don't know how to get started.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi,
I'm very new to speech recognition and pocket sphinx; but having read the tutorials, I'm still not sure how to approach my problem.
My goal is to use a throat microphone with pocketsphinx, but the issue is that the signal outputted by the throat microphone is different from a traditional microphone. In other words, recognition is not very accurate.
I don't know if I need to create a separate language model (even though I'll still be using the same english language) or if I need to try train an acoustic model or just try to adapt the existing english model.
Any insight or intuition would be greatly appreciated. I just don't know how to get started.
You have to train a new acoustic model
You can start with our tutorial http://cmusphinx.github.io/wiki/tutorialam