"William Tracy" <wtr### [at] calpolyedu> wrote in message 
news:48e170f5$1@news.povray.org...
> That gives me an interesting idea: Has anyone tried pairing Povray with a 
> speech-to-text system?

I've spent many hours experimenting with various speech-to-text systems and, 
after hours of training it to recognize your voice and hours training 
yourself to speak in a consistent and clear fashion, it can be made to 
achieve a reasonable level of accuracy where the context enables it to 
rapidly disambiguate alternative interpretations of a word or phrase. Such 
situations include the composition of literary texts where the system can 
use common conjunctions of words to correct potential ambiguities and IVR 
(Interactive Voice Recognition/Response) telephony applications were you can 
present the user with a short menu of verbally distinct options to select 
from.

I'm not aware of any available speech-to-text products that would work at 
all well out-of-the-box for composing POV-Ray SDL. Certain technologies, 
like VoiceXML (designed primarily for IVR) enable you to define voice 
recognition scripts, but it would be a mammoth task to  implement such 
scripts for anything but a very small subset of POV-Ray directives.

Regards,
Chris B.

Post a reply to this message