3 ms·

Or just activate it from known/trained voices?

• khedoros1 10y ago
  I think that recognizing the speaker is a separate, but also difficult, task from recognizing the content of speech.