论文部分内容阅读
A simple and efficient algorithm is presented to separate concurrent speeches. The parameters of mixed speeches are estimated by searching in the neighbor area of given pitches to minimize the error between the original and the synthetic spectrums. The effectiveness of the proposed algorithm to separate close frequencies is demonstrated.
A simple and efficient algorithm is presented to separate concurrent speeches. The effectiveness of the proposed algorithm to separate together speeches. The parameters of mixed speeches are estimated by searching in the neighbor area of given pitches to minimize the error between the original and the synthetic spectrums. is demonstrated.