In this chapter, we introduce hybrid postfilters into speech synthesis, with the objective of enhancing the quality of the synthesized speech. Our approach combines a Wiener filter with deep neural networks. Several ...