EP4082008B1 - Text Independent Speaker-Verification on a Media Operating System Using Deep Learning on Raw Waveforms