----- Original Message -----
From: "Boris Koenig" <[EMAIL PROTECTED]>
To: "FlightGear developers discussions" <[EMAIL PROTECTED]>
Sent: Friday, September 17, 2004 10:22 PM
Subject: Re: [Flightgear-devel] A voice for FG

> John Wojnaroski wrote:
> > Hi,
> >
> > The last month or so I've been working with adding synthetic speech and
> > voice recognition to my 747 project.
> What type of project is that ? - (FlightGear related ?)

See the FG Project webpage for details

Lots of good thoughts, I'll need to digest them over the weekend, but four
quick comments:

1) You could set up a festival server on the Internet, assuming you find the
round-trip delay time acceptable. But just about any older machine might be
faster. I'm using an old AMD-K6 chugging along at 366MHz and find the
performance okay, in fact the delay of a one or two seconds "feels" a lot
like the natural time delay in two-way radio communications

2) It might be just as easy to write a little C/C++ code to define the
interface. Festival uses Scheme so all that is required is a text stream
that mimics a stdin or file input in a Scheme syntax

3) One of the advantages of TTS (at least as I see it) is you don't have to
create snippets or prestore anything. One has the luxury of total and
complete freedom to create any and all possible combinations of words within
the established language model and create the audio in real time.

4) Speech recognition for converting audio to text is just about perfect.
The trick, as you noted,  is having the machine "understand" what was

John W

Flightgear-devel mailing list

Reply via email to