TY  - RPRT
TI  - Improving Trajectory Modelling for DNN-based Speech Synthesis by using Stacked Bottleneck Features and Minimum Generation Error Training
AU  - Zhizheng Wu
AU  - Simon King
PY  - 2016
DO  - 10.1109/taslp.2016.2551865
UR  - https://arxiv.org/abs/1602.06727
ID  - 1602.06727
ER  -