Acoustic Modeling in Statistical Parametric Speech Synthesis - From HMM to LSTM-RNN

   Abstract