Kornel Laskowski, Qin Jin: Modeling Prosody for Speaker Recognition: Why Estimating Pitch May Be a Red Herring. Odyssey 2010: 4