论文部分内容阅读
We propose a mandarin Chinese singing voice synthesis system, in which hidden Markov model(HMM)-based speech synthesis technique is used. A mandarin Chinese singing voice corpus is recorded and musical contextual features are well designed for training. F