Hello everyone,

I've used Sphinx4 to develop a Spanish speech recognition system, using the HUB4 Mexican Broadcast Spanish language & acoustic models.

However, for some reason any special, accented characters, such as ñ and é, are replaced with question marks when Sphinx4 prints the results.

My main class is based off of the LatticeDemo. I've tried changing the output character encoding from ISO-8859-1 to UTF-8, but then it just prints out nonsense characters instead.

I'm using the models and dictionaries from this package.

Here's my configuration file and my main Java class.

Thanks a bunch.

-Eric

--- (Edited on 5/12/2013 9:32 am [GMT-0500] by e.fields) ---