Last modified 9 years ago Last modified on 10/04/07 12:37:44

Welcome to the VoxForge Development Website.

VoxForge was set up to collect transcribed speech for use in Open Source Speech Recognition Engines. We will categorize and make available all submitted audio files and Acoustic Models in GPL format.

The current focus of VoxForge is on collecting transcribed audio for Speech Recognition for IVR (telephony) applications and Command and Control applications on the desktop. We will collect English language audio to begin with, and add other languages in the future.

Click SubmitSpeech on the VoxForge menu above to learn how to record your speech and submit your speech audio files to VoxForge. To learn how 'compile' your speech audio files into Acoustic Models and submit them to VoxForge, click the Create Acoustic Models icon on the Dev menu item above.

For more information on VoxForge click the About link.

Click here to go to VoxForgeDevWiki.


  • April 12, 2007 - The VoxForge Speech Corpus is now separate from the code and scripts used to create the VoxForge Acoustic Models (now located on this site). You can access the Trac Site for the VoxForge Speech Corpus here.
  • April 12, 2007 - You can checkout the source code used to create the VoxForge Acoustic Models using the following command:

$ svn checkout

  • April 12, 2007 - Click here to get to the VoxForge IVR Project - a project that will allow users to submit speech using their telephone (many thanks to trevarthan for coming up with this approach and creating the scripts and Asterisk dial plans)