Is there a simple speech recognition program where you can say a command and that that command will be executed in the terminal? It must be simple to install/setup/configure and it must be a small package because don't have much space left. I am running Lubuntu 14.10 on my EEE PC 701.


Apparently not - there are recognition engines (Sphinx, Julius) but all require a library of speech / meanings - there are several expensive ones out there, and a project to crete an open-source one (which looks to be dead).

Apparently Julius works if you speak Japanese. Sphinx appears perhaps more promising, with updates this year. I don't think it looks simple though.


  • I don't see why not - I'm not a current user of either, so YMMV, but that sort of use is pretty central to these things – Mark Williams May 11 '15 at 12:46
  • The expensive library of speech is misinformation. CMUSphinx provides models for several languages. All available in downloads. – Nikolay Shmyrev May 13 '15 at 22:48
  • Ah - excellent news! I may go back for another look myself then. – Mark Williams May 14 '15 at 13:06

You can use CMUSphinx for that. Tutorial with basics is here.

The simple way to configure this would be to use small Python program. You can find example of voice spotting with Python on CMUSphinx github.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.