LPFT基于局部多项式傅里叶变换的语音盲源分离
关键词 语音信号 局部多项式傅里叶变换 盲源分离 联合近似对角化
毕业论文设计说明书外文摘要
Title Speech Signal Blind Separation Based on Local Polynomical Fourier Transform
Abstract
Currently,the speech signal blind source separation in frequency domain uses STFT and WVD for signal power spectrum. The STFT although is easy to realize and it has no effect of cross terms,but once the window function has been selected,it will cant take into account the time resolution and frequency resolution; WVD method although has high resolution on time-frequency analysis,but it will interfere the separationbecause of the effect of cross terms.So we proposed the method of speech signal blind separation based on local polynomical fourier transform(LPFT),it has less effect on cross terms and can get the information of the frequency more accurately.This paper mainly describes two classical algorithms;studies the basic principles of LPFT,and contrast it with STFT and WVD;analyzes speech signal blind separation based on LPFT,it separating the original siganals effectively with the frequency of the frequency-domain correlation matrix and the separation matrix obtained by the joint approximate diagonalization algorithm.The results in the matlab show that the different speech signals can be successfully separated in the noise environment.
Keywords Speech Signal Local Polynomial Fourier Transform(LPFT)
Blind Source Separation Joint Approximate Diagonalization
目 次
1 引言 1
1.1 研究背景及意义 1
1.2 国内外研究现状 2
1.3 本文主要研究内容 3
2 盲源分离的基本模型和经典算法 5
2.1 线性混合模型 5
2.2 非线性混合模型 8
2.3 盲源分离的典型算法 8
3 局部多项式傅里叶变换（LPFT）盲源分离算法 14
3.1 经典时频分析理论 14
3.2 局部多项式傅里叶变换理论 15
3.3 频域盲源分离模型 17
3.4 局部多项式傅里叶变换盲源分离算法 18
4 仿真实验 20
结 论 23
致 谢 24
参考文献 25
1 引言
1.1 研究背景及意义
随着数字信息时代的来临，网络、通信和计算机技术迅猛发展，数字信号处理这项技术在科技的领域中变得愈来越重要。而盲信号处理，作为现在信号处理领域中研究的热点之一，使广大科学技术人员，对此产生了浓厚的兴趣，也存在着许多潜在的应用。
下一篇：HFSS毫米波微带天线设计