失真测度

本文重点研究语音信号的各种参数表示方法及其相应的失真测度，以及它们在语音质量客观评价方面的应用。

This dissertation focuses on the researches of parameter representations of speech signal and their corresponding distortion measures , as well as their applications to objective assessments of speech quality .

基于DCT子空间失真测度的快速矢量编码算法

A fast encoding algorithm for vector quantization based on DCT subspace distortion measure

VQ话者模型中失真测度的鲁棒性研究

Research on Robustness of Distortion Measurement in VQ Speaker Model

基于信息量失真测度的VQ及在高光谱图像无损压缩中的应用

A VQ Based on Information Distortion Measure and Its Application to Lossless Compression of Hyperspectral Image

使用L1失真测度及其派生方法的无乘法矢量量化

Multiplication-free vector quantization with L_1 distortion measure and the method derived from it

详述了传统码书训练算法LBG的特性以及训练码书的失真测度。

The characteristic of the traditional codebook training algorithm LBG and distortion measurement are also dwelt on .

阐述了Mel谱失真测度的概念，指出用Mel频率尺度可较充分地反映人耳对频率及幅度的非线性感知特性。

The measurement of Mel spectrum distortion is a kind of warped frequency spectrum distortion measure . Using Mel frequency scale can reflect sufficiently the nonlinear perceptive characteristics of human hearings to frequency and amplitude .

提出了一种Bark尺度倒谱系数（BSCC）失真测度。它是采用一种余弦镶边临界带滤波器组代替Mel频率三角滤波器组进行倒谱分析的。

The dissertation proposes BSCC ( Bark-Scale Cepstrum Coefficient ) distortion measure , where a kind of Cosine-fringed Critical Filter-bank , instead of Mel-frequency triangle filter-bank , is employed to frequency warping .

它利用前向神经网络的多维非线性映射原理，将传统的包含平均失真测度计算和平均失真测度到MOS的映射这两个步骤合为一步来实现。

Using the theory of multi-dimension non-linear mapping of the feedfoward neutral network , the method combines the traditional two steps of average distortion computing and mapping from the average distortion value to the Mean Opinion Score ( MOS ) into one step .

经过对不同质量的语音进行仿真实验以及与Itakura测度方法作对比实验，结果表明PSD测度是一种与语音质量主观评价一致性较好的语音失真测度方法。

Preliminary simulative experiments in comparison with the Itakura measure have been done . The results show that the PSD measure is a preferable speech distortion measure that can be consistent with subjective assessment of speech quality well .

数值实验表明，本文提出的增强型巴克谱失真测度（IBSD）与MOS之间具有很强的相关性，能客观地评价出语音信号的主观质量，适用于各种语音编码、语音通信系统。

Numerical experiments show that the improved Bark spectrum distortion ( IBSD ) measurement is highly correlated with MOS . IBSD can estimate the subjective quality of the speech signal . The IBSD can be used in various systems of speech coder and speech communication .

并且通过分析改进型巴克谱失真测度（MBSD）中15个巴克带失真与平均意见分（MOS）间的相关度，提出了一组以相关系数幂次方为权重的有效巴克谱失真权重矢量。

By the analysis of the correlation between each 15 Bark band distortion used in MBSD and mean opinion score ( MOS ), an effective Bark spectral distortion weight vector based on the power of each Bark band correlation coefficient is also presented .

用于客观音质评价的扩展巴克谱失真测度

Extensive Bark Spectral Distortion Measure ment for Objective Speech Quality Assessment

基于感知模型的美尔谱失真测度用于客观音质评价的扩展巴克谱失真测度

Mel-Spectral Distortion Measure Based on Perception Model for Objective Speech Quality Assessment

基于相对失真测度的感兴趣区域编码

An Algorithm for Region of Interest based on Relative Distortion

利用前向掩蔽效应改进巴克谱失真测度

Improvement of Bark Spectral Distortion Measure with Forward Masking Effect

语音质量客观评价更是语音表示及其谱失真测度理论的直接应用。

Objective assessment of speech quality is a direct application of this theory .

基于矢量失真测度和自适应四叉树分割的彩色图象分形压缩方法

Fractal Color Image Compression Based on Vector Distortion Measure and Adaptive Quadtree Partition

基于神经网络的语音谱失真测度研究

Research on speech distortion measure based on neural network

一种基于听觉特性的语音失真测度方法

A speech distortion measure based on auditory properties

在提取声道参数中采用了增益归一化模型谱失真测度〔2〕。

The gain-normalized model of spectral distortion measure [ 2 ] is adopted in picking up vocal tract parameters .

语音信号的表示和谱失真测度一直是贯穿于语音处理各个领域中的重要而基本的问题。

Representation and spectrum distortion measure of speech signals have been a couple of vital issues in research of speech processing .

提出了一种基于相对失真测度的率失真斜率提升感兴趣区域编码算法。

An algorithm based on relative distortion measure for ROI ( Region-Of-Interest ) is proposed which is attained by lifting the rate-distortion slopes .

进一步的实验还表明，该失真测度能提高话者识别系统的时间鲁棒性。

Experiments demonstrate that the distortion measurement presented here can not only improve the speaker recognition system performance , but also enhance its time robustness .

为找出与输入矢量最匹配的码字，传统的穷尽搜索矢量量化编码算法需要计算输入矢量与所有码字之间的失真测度。

To find the nearest codeword for each input vector , traditional full-search vector quantization encoding al-gorithm requires calculating the distance between each codeword and the input vector .

利用前向神经网络，包括多层感知器和径向基函数网络，对多维非线性函数的逼近原理，使得谱失真测度函数具备了表现人耳听觉系统的主观感知行为的能力。

It provide the MSD with the capability of employing subjective behavior of perception of human 's auditory system , by using the theory of approaching to nonlinear function of multilayer perceptrons ( MLP ) and radial basis function networks ( RBFN ) .

谱测度和失真风险测度更具一般性，考虑了投资者风险厌恶对风险测度的影响，其中VaR和ES均为其特例。

When ES ( expected shortfall ) overcomes the pitfall , there are more general techniques such as spectral risk measure and distortion risk measure , in which VaR and ES are both special cases .

描述了利用人耳的听觉掩蔽效应来对巴克谱失真距离测度进行改进，并将这些测度应用在客观音质评价中，以提高主客观音质评价的相关度。

This paper briefly describes a modified method for Bark Spectral Distortion distance measures with the auditory masking effects to improve the measurement relativity and applies the measures to ( objective ) assessments of speech quality .

通过ICA变换得到说话人语音特征基函数系数用于生成VQ码书，并导出包含能量失真的ICA-VQ码书失真测度和质心确定条件，生成最终的判决。

A speaker speech ICA synthesis model was presented to get the speaker speech feature bases with ICA algorithm , and the coefficients of the bases were used in designing codebooks . A novel distortion measurement including energy and a new centroid condition were given .

鉴于不同巴克带谱失真对语音质量的影响程度不同，该算法提出在巴克域中对不同巴克带谱失真进行加权来求失真语音的客观失真测度。

Because each of the Bark band distortion has different effects on the degraded speech quality , this paper proposes the calculation of the objective distortion measurement based on the weighted Bark spectral distortion in the Bark domain .