Page 1 of 5 123 ... LastLast
Results 1 to 10 of 41

Thread: Speech recognition and audio feedback

  1. #1
    Join Date
    Feb 2009
    Location
    Texas
    Beans
    4
    Distro
    Ubuntu 8.10 Intrepid Ibex

    Speech recognition and audio feedback

    I was wondering if there is an embodied conversational agent avalible or would anyone be interested in making one??

    Note: Not like windows's narrator.

    I am talking about a piece of software that will allow its users to communicate and interact with the computer. If it is even possible, which a short google search says it is, does anyone know how to make one?

    PS: I figured that this forum would be a good place to look first.

  2. #2
    Join Date
    Jan 2009
    Beans
    1

    Re: Speech recognition and audio feedback

    I am currently using Dragon NaturallySpeaking 10.0 on Windows Vista 32-bit operating system. Unfortunately this is about the only Speech recognition program available. At one time there was a discussion group working on getting a version of Dragon NaturallySpeaking working in Ubuntu.Apparently the software is closely tied in with Internet Explorer. The researchers were using Wine has an interface between Dragon NaturallySpeaking and Ubuntu

    As you can see I am not a programmer but I do need to use Speech recognition in order to use Ubuntu.

  3. #3
    Join Date
    Aug 2008
    Beans
    302
    Distro
    Kubuntu 11.10 Oneiric Ocelot

    Re: Speech recognition and audio feedback

    There are quite a few speech recognition software(open source software) being developed. Check th wikipedia page on speech recognition in linux.

    I don't know about conversing with the comp though. Sounds like you see sci-fi movies!

  4. #4
    Join Date
    Nov 2005
    Beans
    91

    Re: Speech recognition and audio feedback

    gnome-voice-control
    Is what many have recommended.
    It is based on the Sphinx and PocketSphinx Project.

    site:youtube.com linux voice control
    http://www.youtube.com/watch?v=GCSgkUnlGGA
    http://www.youtube.com/watch?v=w1pyw4fRJH0


    Unfortunately, i haven't had any luck with it.
    Last edited by iiiears; May 14th, 2009 at 02:29 PM. Reason: Added informational links.
    Lawrence Lessig - Copyright in RW/RO culture
    www.youtube.com/watch?v=byB7nKSnFBM Freedom is worth something volunteer your time, talents and dollars. How long will it last? You decide with your support.

  5. #5
    Join Date
    May 2008
    Location
    United Kingdom
    Beans
    4,265
    Distro
    Ubuntu 12.04 Precise Pangolin

    Re: Speech recognition and audio feedback

    The Ubuntu documentation recommends Julius.

    https://help.ubuntu.com/community/Ac...%20Recognition
    Problems with WINE?
    Full Circle Magazine :: Confused "allot"? :: Cheap Linux stickers
    In my day, we had outdoors in which to run, play, and socialise. Now we have computers to do those.

  6. #6
    Join Date
    Jul 2009
    Beans
    2

    Re: Speech recognition and audio feedback

    Trying to use Julius to make a short cut script that runs via wine for pokerstars that also runs via wine so that I can tell it to press zero and it will do the action I assigned zero for in the script on pokerstars. I found a link that helps in showing how to control Rhythmbox. So far having a hard time understanding how to create the vocab files. http://bloc.eurion.net/archives/2008/writing-a-command-and-control-application-with-voice-recognition/

  7. #7
    Join Date
    Feb 2007
    Beans
    Hidden!

    Re: Speech recognition and audio feedback

    If you want to ask questions about Julius, check out the forums at VoxForge first. VoxForge is dedicated to bringing good-quality open source speech recognition to the community and there is a lot of know-how available over there.

    I have also contributed a lot of speech and I think so should everyone else who is interested in open source speech recognition.

  8. #8
    Join Date
    Oct 2006
    Location
    Ohio, USA
    Beans
    1,775
    Distro
    Ubuntu 10.04 Lucid Lynx

    Re: Speech recognition and audio feedback

    Quote Originally Posted by rvk View Post
    If you want to ask questions about Julius, check out the forums at VoxForge first. VoxForge is dedicated to bringing good-quality open source speech recognition to the community and there is a lot of know-how available over there.

    I have also contributed a lot of speech and I think so should everyone else who is interested in open source speech recognition.
    What does that mean, donate speech? Record your voice for them? And how do you do it?

  9. #9
    Join Date
    Nov 2009
    Beans
    26
    Distro
    Ubuntu 9.10 Karmic Koala

    Re: Speech recognition and audio feedback

    There IS an AI bot called "ainebot". The personality I spoke with is named Amy but there are others, and I snarfed it from the Puppy forum. I didn't use the Voice recognition with it yet. LEGAL DISCLAIMER (to satisfy the lawyers I don't actually have) each instance of "talk" should be regarded as a typed text string... (like I'm talking to you right now)

    Amy is a bit difficult to talk to on the subject of computers, every instance of the term "computer" is met with "I have an IBM Thinkpad, it's CUTE!"

    BUT some of her questions regarding (yes, she asks questions rather than just respond...) the differences between robots and Other Humans show that she's actually aware that she's not the same kind of human we are. (most of us are) How aware I don't know. It's kind of distracting trying to hold a conversation with her. She'll get to conversing and just when you think you're understanding, and being understood, she'll take it All The Way Left. Pretty damned impressive in any case.

    One of the other personalities, CharliX I believe, can actually write scripts. It's near the Holy Grail of geekdom, a computer you can tell to program itself.

    As for conversing using Speech Recognition there's a WinBloze app called Virtual Hypnotist, so far not ported to *nix but it was written in Visual Basic, thus depends on a lot of Microsux products. Sapi4 and Sapi5 w/ M$ Dictator. Although the SR part that you can actually converse with only works with Sapi5 and you're limited to the Voices you get with it.

    There's a Mac app called Dictate that does the same thing Winbloze does, just better, and it's Unix based, so a quick kludge might work. Don't know yet.

    The virtual hypnotist app lets you design your own hypnosis sessions, with the help of bots.

    All that off the Science Fiction post.

    There's another set of applications I'd like to see, far better, I'd actually like to have the skills to hack it together myself, Alas, I don't.

    But one of the goals is to have Linux or indeed ANY OS and especially the collection of Ideas and Ideals I like to call "an Internet".( You can feel free to use the term, I never had it copyrighted. )

    But to have the Web accessible, fully, to blind and otherwise physically disabled persons.

    Which is where I came into this thread.

    My communications skills are limited by a neuro condition called Aspergers. Talking to me is a lot like talking to Amy. Writing with pen and paper is a nightmare for me and worse for anybody who has to try to read it.

    But I can type wildly fast and accurately and between forums and chat can actually communicate. It feels great.

    And it got me intrigued about a lot of communications barriers.

    Meanwhile if one or more of you want to try a kludge between voice recognition, speech synth and Ainebot please do. I haven't had much success with it, yet, but then, I'm only one person out of 6 billion.
    "Damn, that was the stupidest deal I ever made"
    Odin, immediately after he traded his eye for wisdom.

  10. #10
    Join Date
    Nov 2009
    Beans
    26
    Distro
    Ubuntu 9.10 Karmic Koala

    Re: Speech recognition and audio feedback

    The guy who developed the Virtual Hypnotist suite of apps: followthewatch55@yahoo.com

    I tried getting the Microsux speech API to run in Wine but I might not have the right hardware for it. Eventually I'll get to the point where I can hack software the way I hack hardware now.

    For now, one of y'all might be able to do something.

    There's another Alternative Input device I have downloaded, where you use your webcam as a mouse . Two of them gxneur and MouseTrap. Combine that with a pen-and-paper device from the 1890s called The Chevreul Pendulum. The possibilities come together.

    gxneur is a translator program for keystrokes. Between that, speex and Dictate... AND... MouseTrap uses the camera as a mouse, and the Pendulum, it traces through a piece of string and a needle and a 6-way pattern, neuro-muscular impulses that communicate with the simple words "yes, no, maybe, not sure and don't want to answer". Google it.
    "Damn, that was the stupidest deal I ever made"
    Odin, immediately after he traded his eye for wisdom.

Page 1 of 5 123 ... LastLast

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •