Nate Nichols wrote:
Hi,
I was wondering if anyone knew how the automatic phoneme extraction in
FacePoser works. I always kind of assumed that it was an easy SAPI
call, but after poking around through the SAPI documentation, I'm
starting to think this is something Valve wrote themselves.
Does anyone know what the phoneme extractor is actually doing? I
assume it's some kind of sliding window of recognition using the text
you give it, but if anyone has any more specific thoughts or ideas I
would love to hear them.
Or, better yet, is Valve considering opening any of this code up?
My guess is that they are doing an untrained dictation mode transcription.
You can find several examples of SAPI code on the net, for example...
http://www.ccir.ed.ac.uk/~jad/phonemes.html
--
Jeffrey "botman" Broome
_______________________________________________
To unsubscribe, edit your list preferences, or view the list archives, please
visit:
http://list.valvesoftware.com/mailman/listinfo/hlcoders
