These are the programs implementing the actual speech recognition:
There are several packages for driving Emacs through voice input, including:
Allows use of DragonDictate from an X window system. http://www.cl.cam.ac.uk/a2x-voice/a2x-faq.html
VoiceCode is an Emacs interface to Dragon NaturallySpeaking (uses Natlink). https://sourceforge.net/projects/voicecode/ The last update to the SVN repository was in 2013.
A voice-controlled keyboard for GNU/Linux built on sphinx. http://sourceforge.net/projects/voicekey/ (unmaintained since August 2013, see https://sourceforge.net/p/voicekey/voicekey/ci/329cc8daac480349a3308121d9e8d35a01725ed4/)
Emacs interface to Dragon NaturallySpeaking. http://emacs-vr-mode.sourceforge.net/
The last release announcement (as of Oct 2013) was in 2007… http://sourceforge.net/mailarchive/forum.php?forum_name=emacs-vr-mode-announce
EmacsListen is an implemention of ShortTalk. It is based on vr-mode.
http://shorttalk-emacs.sourceforge.net/
EmacsListen is unmaintained. The latest commit in its CVS Repository was in 2004. – XueFuqiao