Talk:Speech recognition software for Linux

In case the admin's here at Wikipedia start deleting things, you might check for the most recent version in Deletionpedia and move them to my wiki or Wikinfo. These wikis allow content that is based on original research and self-published sources. Lumenos (talk) 23:07, 15 November 2008 (UTC) This article only links from one important page "speech recognition" so it would not be difficult to relink that page to Wikinfo or Lumeniki. Lumenos (talk) 23:31, 6 March 2009 (UTC)

Article title would probably more correctly be 'Speech recognition on POSIX', there being nothing specific to the popular kernel, however that wording would be ugly. Verte ttg (talk) 12:29, 2 July 2008 (UTC)
 * POSIX is not ugly it just doesn't have the name recognition. And besides it is GNU/POSIX. (kidding) Lumenos (talk) 02:30, 4 September 2008 (UTC)
 * I think the name change is a good idea, but I think the link(s) should say "Linux" due to its popularity.Lumenos (talk) 23:31, 6 March 2009 (UTC)

Perhaps the trouble is that the article is trying to cover both Free speech recognition software, and hacks to get other speech recognition software working in a virtualised environment or with wine. Verte ttg (talk) 12:29, 2 July 2008 (UTC)
 * Virtualization is not what I would call a "hack", maybe DNS under Wine or WinDictator is. But I don't see the problem with having one article for all this, so long as it is not too long. If you want to know what speech recognition is possible with Linux, it shows you the options. Lumenos (talk) 02:30, 4 September 2008 (UTC)

Proposed addition: LumenVox Speech Engine
Though a closed-source and commercial application, I would suggest that my employer's product, the LumenVox Speech Engine, is worthy of a mention on this page. We build it and run it on Linux (with support for RHEL/CentOS, Fedora Core, and Debian) and it provides a complete speaker independent solution on Linux.

Note that it is not dictation software, so it is not a substitute for a product like Naturally Speaking. It is a tool for developers to build applications around, and the API is fully documented online. Though it is aimed mainly at developers building telephone systems (IVR applications), it can be used for any application that needs to recognize short utterances from a variety of speakers, e.g. command-and-control applications.

It is also integrated with Asterisk, a popular open source PBX run mostly on Linux, which may make the product more noteworthy for people looking for information about Linux speech recognition.

If any editors are interested in making this addition, I can provide more information on the product, including more references.

Full disclosure: I am employed by LumenVox, which sells a commercial automatic speech recognizer. I make this disclosure in compliance with the WP:SCOIC guidelines. While I am a longtime user of Wikipedia, I have no experience as a contributor and would appreciate any help or guidance from editors. Stephen Keller (talk) 22:40, 10 March 2009 (UTC)
 * I've added LumenVox as it has plenty of references on the web and is supported by the Digium, the Asterisk people. While I was at it I sorted and expanded existing list entries by adding snippets explaining what some of the bare links are. "" template removed Kiore (talk) 21:22, 4 April 2009 (UTC)
 * Looks great, Kiore. One minor typo: In the sentence, "It has been integrated into the Asterisk," I think you have the word "the" in there where it doesn't belong. Stephen Keller (talk) 16:57, 6 April 2009 (UTC)
 * Thanks Stephen, sloppy proof reading on my part. I intended the "the", but forgot to add the rest of the sentence "into the Asterisk private branch exchange system". Now fixed.Kiore (talk) 19:29, 6 April 2009 (UTC)

The introduction
I'm wondering if the term 'software packages' in the intro needs some expansion in the general context of the intro. It could imply a single comprehensive utility that does all of TTS, speech recognition and dialog manager (e.g. Simon). However it is possible to combine several individual 'software packages' to provide an end-to-end solution (user speaks a command, computer acts on that command, perhaps with verbal response) in the speaker-dependent domain with sufficient accuracy to make it useful. Consider CMU Sphinx or alternately Festival+HTK+Julius plus a Python dialog manager. It is something of an individual mashup in the open source domain but perhaps the article might point out that this is already possible.Colbec (talk) 16:52, 8 April 2012 (UTC)

Offer to edit the page from ground up
I am very interested in this subject matter of voice and speech recognition and would like begin to edit the reference section. As I am a new editor, would somebody like to team up with me ? — Preceding unsigned comment added by Alwerner (talk • contribs) 18:32, 28 February 2013 (UTC)

Out of date
"This page was last edited on 10 April 2014, at 12:45." Ststeve11 (talk) 03:30, 29 April 2017 (UTC)

This article is not too technical, just need to restructure and update
The title say that this article is about speech recognition software for Linux. It should analyze the speech recognition task, in order to have a framework for comparison of all the known programs written for the Linux OS. Then a description of each program, finishing with a comparison table. Of course with links to articles about each program, and the corresponding links to each project. I am not sure if windows programs that can run with wine should be included, but I am sure that Windows programs running in a windows virtual machine, should be excluded, because they are not running in Linux, they run in other machine. I tried to fix the section about the speech corpus used for training the recognizers. I think that the speech corpora used to train each recognizer should be mentioned as a feature of each one. That part should be removed or if it is important include a section mentioning each available corpus mentioned in the comparison. — Preceding unsigned comment added by 2806:107E:C:F09:218:DEFF:FE2B:1215 (talk) 07:25, 21 September 2018 (UTC)

Page should be deleted
This page seems completely outdated and redundant.

It might seem bold but I'm convinced it's better than greeting visitors with several warning signs the content is outdated, links are not trusted etc.

It reflects badly. (Disagree. This is not a reason to remove this info. As a user, not an editor I was really glad to find this helpful article. Wikipedia is about info not vanity... Tom Midgley )

Please someone who is versed in Wiki-ing, take the bold step, regards — Preceding unsigned comment added by 195.176.238.195 (talk) 11:50, 29 May 2019 (UTC)