Multi-output RNN-LSTM for multiple speaker speech synthesis with α-interpolation model | Publicación