A HMM-based speech synthesis system using a new glottal source and vocal-tract separation method.

Pierre Lanchantin, Gilles Degottex, Xavier Rodet

Browse the full ICASSP paper archive.