Natural text-to-speech and voice recognition?

[GimbusRictumFlapper]

What is the current state of freeware text-to-speech
(preferably the natural sounding kind) and voice
recognition software? Particularly in relation to
open source development?

 

[B. R. 'BeAr' Ederson]

What is the current state of freeware text-to-speech
(preferably the natural sounding kind) and voice
recognition software? Particularly in relation to
open source development?

For TTS I liked the combination of MBROLA with Euler:

http://www.tcts.fpms.ac.be/synthesis/euler
http://tcts.fpms.ac.be/synthesis/mbrola.html

But I haven't used it for a couple of years. So I can't say anything
about the current state of these packages. I used the free engine of
L&H TTS, too. But it didn't produce a real natural sounding output,
although you could customize pauses and intonation. I don't believe
that L&H have released a newer version for free. So it maybe isn't
worth trying again.

A quick web search on the above mentioned tools showed up this page:

http://sig.levillage.org/index.php?p=551

It seems worth reading.

I can't say anything about voice recognition. Somebody else will sure
help you with this. ;-)

BeAr