2kb/s Waveform Interpolation Speech Coding Based on Non-negative Matrix Factorization

ZHANG Peng; BAO Chang-chun; GUO Li-li

您当前的位置：

首页 >

文章列表页 >

2kb/s Waveform Interpolation Speech Coding Based on Non-negative Matrix Factorization

更新时间：2025-07-16

- 2kb/s Waveform Interpolation Speech Coding Based on Non-negative Matrix Factorization
- Acta Electronica Sinica Vol. 36, Issue 4, Pages: 632-638(2008)
- 作者机构：
  
  北京工业大学电子信息与控制工程学院,北京,100022
- 作者简介：
- 基金信息：
- DOI：
  CLC： TN912.3
- Published Online：25 April 2008，
  
  Published：2008
- 稿件说明：
移动端阅览
ZHANG Peng, BAO Chang-chun, GUO Li-li. 2kb/s Waveform Interpolation Speech Coding Based on Non-negative Matrix Factorization[J]. Acta Electronica Sinica, 2008, 36(4): 632-638.
DOI：

ZHANG Peng, BAO Chang-chun, GUO Li-li. 2kb/s Waveform Interpolation Speech Coding Based on Non-negative Matrix Factorization[J]. Acta Electronica Sinica, 2008, 36(4): 632-638. DOI：

摘要

在波形内插(Waveform Interpolation

WI)语音编码器中

如何低延时、高精度并且低复杂度的分解和量化特征波形(Characteristic Waveform

CW)一直是该编码模型的研究热点和难点.本文提出用非负矩阵分解(Non-negative Matrix Factorization

NMF)方法来分解语音特征波形.该分解方法仅需要当前帧的语音信号

不会给编码器带来额外的延时;为了提高分解精度

本文在CW分解之前先对CW按照其子帧的最大基音周期进行分类

然后按不同类别进行分解.另外

本文结合耳蜗模型提出了NMF的基矢量分带初始化算法

将CW的分解精度提高到与二阶奇异值分解相当的水平;为了降低WI编码器的计算复杂度

本文去除了传统WI编码器中的特征波形对齐模块

同时将NMF的分解阶数设定为16以折中CW分解的计算复杂度和分解精度.最后

本文基于矩阵量化技术

对非负矩阵分解后的编码矩阵采用分裂式矩阵量化方案来量化.主观A/B测试表明

本文提出的2kb/s NMF-WI编码器的合成语音质量接近于2.4kb/s SVD-WI编码器.MOS分测试表明

本文提出的2kb/s NMF-WI编码器的合成语音质量稍差于2.4kb/s MELP编码器.

Abstract

In WI coding scheme

how to decompose and quantize the characteristic waveforms with low delay

low complexity and high precision have always been a hot research topic.The characteristic waveform decomposition based on non-negative matrix factorization is proposed in this paper.This CW decomposition method doesn't bring any additional delay to WI coder;In order to improve decomposition precision

the CW is firstly classified according to the maximum pitch of its sub-frames before being decomposed.Besides

band-partitioning initialization constraints are set to basis vectors before NMF is carried out

and this has made the CW decomposition precision of NMF-based method be comparable with that of 2 ranks of SVD;In order to reduce the computational complexity of WI coder

the CW alignment procedure is removed in our NMF-WI coder

and the factorization rank of NMF is set to 16 as a trade-off between computational complexity and decomposition precision.In the end

the low dimensional coding matrix is quantized by splitting matrix quantization scheme.The subjective A/B listening tests show that the proposed 2kb/s NMF-WI coder can give smooth speech with quality close to 2.4kb/s SVD-based WI coder.Mean Opinion Score test results indicate that the performance of proposed coder is a little worse than that of 2.4kbps MELP coder.

关键词

Keywords

references

Views

1292

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

2kb/s Bayesian Ying-Yang Waveform Interpolative Speech Coding Based on Non-Negative Matrix Factorization

1kb/s Waveform Interpolative Speech Coding Based on Two-Dimensional Nonnegative Matrix Factorization

Waveform Interpolation Speech Coding Based on DCT

Low Bit Rates Waveform Interpolation Speech Coding Based on Singular Value Decomposition

Super Resolution Pitch Detection Based on Band-Partitioning Spectral Entropy and Signal Decomposition in DCT Domain

Related Author

BAO Chang-chun

GUO Li-li

FONT

LI Ru-wei

BAO Chang-chun

XUE Er-juan

Verdana

FONT face

Related Institution

Speech and Audio Signal Processing Laboratory, School of Electronic Information and Control Engineering, Beijing University of Technology

Speech and Audio Signal Processing Lab, School of Electronic Information and Control Engineering, Beijing University of Technology

School of Electronic Information and Control Engineering Beijing University of Technology

100084

清华大学微电子学研究所!北京

⁰