Factor analyzed voice models for HMM-based speech synthesis | IEEE Conference Publication | IEEE Xplore