叠加特征信息辅助的语音传输与重构

Speech Transmission and Reconstruction Assisted by Superimposed Feature Information

为改善压缩语音传输系统的重构精度且不增加系统的频谱开销,提出一种叠加特征信息辅助的语音压缩传输与重构方法。提出方法首先提取稀疏语音信号的特征信息;抽取的特征信息以叠加序列方式叠加在压缩语音信号上进行传输;接收机重构时,借助特征信息辅助重构算法进行语音重构。分析与仿真结果表明,相比于传统的压缩感知语音重构方法,在较高信噪比或较低压缩率情况下,提出方法可改善语音重构精度,且不增加传输系统的频谱开销。

To improve the reconstruction accuracy of compressed speech transmission system without increasing the spectrum resource overhead, a method of speech compression transmission and reconstruction assisted by superimposed feature information is proposed in this paper. The feature information is extracted from the sparse speech signal. The extracted feature information is superimposed on the compressed speech signal for transmission. At the receiver, the superimposed feature information is recovered, which is employed to assist the reconstruction algorithm to reconstruct the speech signal. Compared with the traditional compressed sensing-based speech reconstruction method at higher signal-tonoise ratio or lower compression ratio, the analysis and simulation results show that the proposed method can improve the speech reconstruction accuracy without increasing the spectrum resource overhead of the transmission system.