I would want to extend the question to include cmuclmtk, and the sphinx set of
tools in general. What are the things that would need to be done differently
for a right to left language?
Feedback highly appreciated.
Thanks.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hello folks,
If one were to develop a speech recognition system for a right to left
language (e.g. Arabic, Persian etc), would the sphinxtrain tools support that?
Please share your thoughts.
I would want to extend the question to include cmuclmtk, and the sphinx set of
tools in general. What are the things that would need to be done differently
for a right to left language?
Feedback highly appreciated.
Thanks.
People,
Any thoughts?
Everything is the same. Right-to-left concerns only of display of the words,
not to the order of them.
it does not make any deference , the problem you will face if you want to
display the results at windows platform you need to deal with Unicode issues
I have also a similar issue, should the words in sentences of training text be in reverse order for statitical model and accoustic model training?