Hybrid Quantization Line Spectrum Pair(LSP) Vocoder For various emission bands the coefficient b has different values. 混合量化线谱对(LSP)声码器对于不同的谱线,系数b是不同的。
Line Spectrum Pair(LSP) is another form with LPC, belong to the frequency parameter. 线谱对(LSP)是线性预测参数的另一种表示形式,同样属于频域参数,可以用来估计基本的语音特性。
First this paper introduces the principle of ACELP, then the relation between linear prediction ( LP ) coefficient and line spectrum pair ( LSP ) is provided. 首先本文介绍了ACELP编码的基本原理,然后给出了线性预测(LP)系数与线谱对(LSP)(LSP)之间的转化关系。
A semi-fragile speech watermarking algorithm based on speech model is researched, which is realized by quantization of line spectrum pair ( LSP ) parameters. 在语音产生模型的基础上研究了一种基于该模型的半脆弱性数字水印算法,通过对语音信号的线谱对(LSP)(LSP)参数嵌入水印信息来实现。
After researched with LPC algorithm, A Line Spectrum Pair(LSP) ( LSP ) Based Algorithm for Helium Speech Enhancement is proposed in this paper. 所以,本文将频域的线谱对(LSP)(LSP)分析应用于氦语音增强中,提出了一种基于线谱对(LSP)系数(LSP)的氦语音增强算法。