3 ms·But matching a speaker to a face is a different problem differentiating between speech in a recordingby totetsu 4y agoBut matching a speaker to a face is a different problem differentiating between speech in a recording