2 ms·High-Quality Deep Fake audio and video with only 40 seconds sample training time14 points by fortran77 6y ago