Xin Wang, Shinji Takaki, Junichi Yamagishi: A Comparative Study of the Performance of HMM, DNN, and RNN based Speech Synthesis Systems Trained on Very Large Speaker-Dependent Corpora. SSW 2016: 118-121