.Net framework inlcudes the superb speech recognition and speech synthesis module.
Idea here: A link between a .Net module and the client patch.
Written texts (console etc., instructions) can be heard as voice output on the client. Simply needs the introduction of a single command available from scripting.
Speech recognition works the other way round, recognized speech is injected into the clients engine (like stufftexting maybe) or available to scripting somehow (would make it possible to control parts of the game/gameplay via voice)
Microsoft windows comes shipped with default voices for speech synthesis, the "old" voice is called "Microsoft Sam" which is the voice available in windows XP.
Vista and Win7 is shipped per default with the new voice "Microsoft Anna"
The capability and quality of the spoken text depends much on this voice. Sam + Anna are "english speakers", on vista + win7 is also the chinese voice "L" (dont remember the exact name atm) available. There are many other voices available from 3rd party developers (not free) like arabic etc.
Speech synthesis is simple. Any text can be converted to speech. Speech recognition is bit more difficult, it relies on doing some training so that the spoken words are properly recognized, but once trained its 99% accurate.
I have attached a little demo (only speech synthesis) to this post, extract the rar archive and run the .exe (needs .net framework)
P.S.
There is a trick on how to make the new voice "Anna" available to windows xp users, (and there is no other way as I know of):
Download Streets + Trips Trial from microsoft (1.2 GB): http://www.microsoft.com/streets/en-us/free-trial.aspx
You can either install the complete package then you will have "Anna" on your computer or you can cancel the installation after the voice has been installed.
Incorporation of Speech recognition and synthesis
Reply With Quote