Support for Mac / Linux

[GoogleCodeExporter]

Are you planning on building support for Mac or Linux?  I'd would be so awesome 
if this module 
was cross-platform!

Original issue reported on code.google.com by kyle.fox on 15 Aug 2008 at 5:27

[GoogleCodeExporter]

Hi Kyle,

Unfortunately, not Linux -- nobody has written a speech engine for Linux yet 
that I'm
aware of.  I don't own a Mac to do the necessary research into how to hook into 
a
Mac's Speech engine, but I would absolutely love patches! ;)

Michael

Original comment by gundl...@gmail.com on 15 Aug 2008 at 10:19

• Changed state: Accepted

[GoogleCodeExporter]

There are several projects devoted to speech synthesis and recognition on Linux.

This works pretty well for TTS:
http://www.cstr.ed.ac.uk/projects/festival/

I got all these from:
http://en.wikipedia.org/wiki/Linux_speech_recognition_software

http://cmusphinx.sourceforge.net/html/cmusphinx.php
http://julius.sourceforge.jp/en_index.php
http://freespeech.sourceforge.net/
http://live.gnome.org/GnomeVoiceControl

This is the microsoft algorithm used to learn speech.  It is free, but cannot be
redistributed: http://htk.eng.cam.ac.uk/ It works with Linux.

IBM ViaVoice used to be free, and was the best recognition engine that I have 
used,
but has been discontinued.

Also, it is possible to use programs such as Dragon NaturallySpeaking 9 in 
Linux by
utilizing Wine, though some problems will arise

Original comment by david.ko...@gmail.com on 3 Sep 2008 at 12:29

[GoogleCodeExporter]

As for mac, it's pretty easy to say stuff:
http://www.mactips.org/archives/2005/07/29/command-line-text-to-speech/

I think the speech recognition stuff requires some apple-scripting, but I could 
be
wrong. Here is a cool video:
http://lifehacker.com/software/speech-recognition/hack-attack-make-your-macs-spe
ech-recognition-work-for-you-215764.php

Original comment by david.ko...@gmail.com on 3 Sep 2008 at 12:33

[GoogleCodeExporter]

Hi David, thanks for all of the data!  I don't care too much about getting TTS 
to
work, as pyTTS is really the package to use for precise voice synthesis; speech
recognition is what this package is really about.

Again, as I don't have a Mac, I'd appreciate anybody who could send me a Python
snippet that registered a callback for speech events.

As for Linux, when I did my research a few weeks back I got the impression that
Sphinx was nowhere near ready for prime time.  It looks like I'll have to do 
more
digging :)

Original comment by gundl...@gmail.com on 3 Sep 2008 at 12:14