论文部分内容阅读
This paper presents an improved voice morphing algorithm based on Gaussian Mixture Model(GMM) which overcomes the traditional one in the terms of overly smoothed problems of the converted spectral and discontinuities between frames.Firstly, a maximum like