Using voice-quality measurements with prosodic and spectral features for speaker diarization | Publicación