A Cooperative Voice Analysis Repository for Speech Technologies






We believe that the COVAREP repository has a great potential benefit to the speech research community and we hope that you will consider contributing your published algorithms to it. Please see below a few instructions for this purpose.

Submission process

  1. Discuss the Intellectual Property (IP) issues with the copyright owner of your code and pick up a convenient License which can be linked to the General Public License Version 3 (GPLv3) (see diagram at License Compatibility or to help you chose a license ). We would encourage you to try to obtain the LGPL licence from your institution in order to be more industry friendly.
  2. Update your code to minimize redundancy with the functions already present in the last version of COVAREP and the Voicebox.
  3. Update your code to make it comply with the Coding Convention while using the Function Template
  4. Make a HOWTO script file (see the howtos directory in the repository for examples) which exemplify your code.
  5. Test your code on many platform as possible (at least one Linux, Mac or Windows). Note that platform specific code is not accepted.
  6. Attach at least one already published article which describes the suggested method in details. Because we want to ensure a minimal scientific quality and relevance, the article has to be published in one of these International Conferences or Journals.
  7. Send your submission either through the GitHub website or by sending the code and the paper by e-mail to one of the current maintainer listed below.
  8. A maintainer of the COVAREP repository will handle your submission and ask for modification if any points above has to be clarified.

Current maintainers

List of accepted International Conferences and Journals

If the conference or journal of your publication does not appear in this list, please send an e-mail to the current maintainers to ask them to consider adding it: Gilles Degottex <>, John Kane <>, Thomas Drugman <>, Tuomo Raitio <>, Stefan Scherer <>