HMM-based speech synthesiser using the LF-model of the glottal source | IEEE Conference Publication | IEEE Xplore