I am trying to build a speech recognition model. This model should be trained on audio data to learn how sentences are made and how words are pronounced. The problem is most audio files (speeches, audiobooks, interviews..) are GPL/LGPL licensed. Should my work be published under GPL too? Is my model considered a derivative work of the audio data?
PS: 1- The data is used only for training and will not be distributed in any way
2- The data can't be reconstructed from model

  • 1
    Which versions of the licenses are the data files licensed under exactly? GPL/LPGL are for software. I would expect audio to be licensed under OpenAL or Creative Commons or something similar. – Prim Reaper Sep 8 '20 at 15:34
  • Most are CC BY-NC-SA although some websites claim GPL even for audio data. – mobelahcen Sep 9 '20 at 7:16

Your Answer

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

Browse other questions tagged or ask your own question.