一种新的规则脉冲激励语音合成模型

A New Regular Pulse Excitation Voice Synthesis Model

本文提出了一种新的规则脉冲激励(RPE)合成模型.它是在原 RPE 模型的基础上融合了清浊音模型对清音的处理手法而得到的.它所合成的语音比原RPE 模型有所提高.

A new regular pulse excitation synthesis model is presented.On the basis of the original RPE model,this model benefits from the method of voice/ unvoiee model processing the unvoice speech.In the synthesis terminal,the excitation to the synthesis filter has white noise besides the regular pulse ex- citation.Compared with the original RPE model,this new model produces better speech.