17

Anyone have a simple tutorial for running speech recognition under linux? I see that pocketsphinx is available as a binary download in the software centre, but running it from terminal fails reporting that it needs parameters, but I do not know what to put there. I tried installing all the voice models listed, but I don't know where they get saved or how to make them work. I also wasn't sure what to do to make Sphinx3 work either. The Sphinx3 manpages seem broken and the pocketsphinx one doesn't have an example usage that makes sense to me.

I'm looking for something that gets 60% correct recognition from microphone input with a limited vocabulary and words, and can then write that to a text file - this stuff existed in the late 1980s. So, I know it exists and should work somehow.

Thanks.

(P.S. Internet connection is a no go)

user2068060
  • 315
  • 1
  • 3
  • 8
  • I had trouble getting http://www.simon-listens.org running on Ubuntu last time I tried, but that was quite a while ago, might be worth investigating – lofidevops May 20 '13 at 12:01
  • 1
    You might want to stipulate whether requiring an active internet connection is okay. The google voice service accepts an audio file and returns text, but it's not offline capable. – RobotHumans May 20 '13 at 12:47
  • Active internet is a no go. – user2068060 May 20 '13 at 15:51
  • 1
    The best way to add additional information to your question is by editing it, with the *edit* button. It is better visible that way, and comments are mainly for secondary, temporary purposes. Comments are removed under a variety of circumstances. Anything important to your question should be in the question itself. – guntbert May 20 '13 at 19:35
  • Related: https://askubuntu.com/questions/161515/speech-recognition-app-to-convert-mp3-to-text – Ciro Santilli OurBigBook.com Oct 07 '20 at 16:49
  • 1
    Does this answer your question? [Speech-recognition app to convert MP3 to text](https://askubuntu.com/questions/161515/speech-recognition-app-to-convert-mp3-to-text) – Flimm Aug 23 '23 at 13:26

3 Answers3

4

Try Simon Listens

Simon is an open-source speech recognition program and replaces the mouse and keyboard. Its designed to be very flexible and allows customization for any application where speech recognition is needed.

To Download and more information visit simon on kde.org

Mark Kirby
  • 18,289
  • 19
  • 78
  • 113
Mitch
  • 106,657
  • 24
  • 210
  • 268
1

Try to use speech to text in linux, based on google speech recognition and integrated in Linux applications

Алексей Шечков
  • 11
  • 1
0

The accuracy rate depends on multiple factors such as accent, background noises, quality of recordings. The pre-trained models are a bit limited, and used Transcribear instead a web browser based speech-to-text tool in Linux for my transcription projects.

John
  • 71
  • 3