A comparison of open-source segmentation architectures for dealing with imperfect data from the media in speech synthesis.

Ascensin Gallardo-Antoln, Juan Manuel Montero, Simon King

Browse the full Interspeech paper archive.