Towards an Interpretable Representation of Speaker Identity via Perceptual Voice Qualities

Open in new window