Corpus Download Link


I have 2 questions…

First: Is there a link to download always the newest release of a Common Voice Corpus for a specific language?

If so, it would be easy to download the corpus through an audio corpus tool like audiomate. Adding the download possibility to such tools would increase the usage of the dataset, so should be possible.

When the following is possible, whats about the email and the “I Accept I will not try to recognize persons in the dataset” checkbox? Can it just be omitted, or is it a MUST to ask, maybe only the “Accept” Question, for these 2 things?

Maybe the real second question is: Should the dataset be downloaded through the Common Voice website or is it OK to download the dataset through former mentioned tools (e.g. audiomate).


1 Like

For legal reasons (ability of contributors to request removal of their data due to GDPR) the dataset must be downloaded through the common voice website and checking the checkmarks there.