I'm using PocketSphinx on android and everything is going well except for one thing now.
I created my dictionary and langage model using this: http://www.speech.cs.cmu.edu/tools/product/1463688616_24344/
It works very well and even a bit too well; when I say a word included in the langage model, it is recognized accurately, however if I say other things, it is somehow recognized aswell as word from my lm.
For instance, lets say my lm contains "house forest yellow " and I say "glasses", it will recognize "house" (I didn't try with those specific words, it is just to illustrate the issue.)
Is there a way to avoid that ?
Or another way to do what I want ? I think I had a similar issue using the grammar based search.
I thought about using a keyword activation search to limit the problem a bit but I don't think it is enough.
Thank you in advance,
Terry
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hello,
First, thank you for your awesome work.
I'm using PocketSphinx on android and everything is going well except for one thing now.
I created my dictionary and langage model using this:
http://www.speech.cs.cmu.edu/tools/product/1463688616_24344/
It works very well and even a bit too well; when I say a word included in the langage model, it is recognized accurately, however if I say other things, it is somehow recognized aswell as word from my lm.
For instance, lets say my lm contains "house forest yellow " and I say "glasses", it will recognize "house" (I didn't try with those specific words, it is just to illustrate the issue.)
Is there a way to avoid that ?
Or another way to do what I want ? I think I had a similar issue using the grammar based search.
I thought about using a keyword activation search to limit the problem a bit but I don't think it is enough.
Thank you in advance,
Terry
You need to use keyword spotting