I am trying to develop a native speech recognition for my thesis where it is only limited in recognizing a word (a cebuano word specifically) and I will be using CMU Sphinx.
How can i create the language model if i only want to recognize a single word?
And what kind of recordings should i record if this is the case? Thanks a lot!
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Just one word does not make it easier. Record your langauge and transcribe it. Record yourself and your relatives, record stories and then share them with transcription. Start with 5 hours of recordings of 5-10 people. Include the important words of interest there as well. Then build the models as suggested in tutorial
I am trying to develop a native speech recognition for my thesis where it is only limited in recognizing a word (a cebuano word specifically) and I will be using CMU Sphinx.
How can i create the language model if i only want to recognize a single word?
And what kind of recordings should i record if this is the case? Thanks a lot!
Just one word does not make it easier. Record your langauge and transcribe it. Record yourself and your relatives, record stories and then share them with transcription. Start with 5 hours of recordings of 5-10 people. Include the important words of interest there as well. Then build the models as suggested in tutorial
http://cmusphinx.sourceforge.net/wiki/tutorial
Last edit: Nickolay V. Shmyrev 2016-01-24
contact mo na lang ako dai. Help kita. email mo ko sa leimiaoren@gmail.com
Thanks for the immediate response sir! I have another question, will the 5-10 people have the same recordings?
Different ones