Weakly-supervised text-to-speech alignment confidence measure.

Guillaume Serrire, Christophe Cerisara, Dominique Fohr, Odile Mella

Browse the full COLING paper archive.