Excitation-by-SampleRNN Model for Text-to-Speech | IEEE Conference Publication | IEEE Xplore