Voice conversion is an newly branch of speech signal processing.
说话人转换是语音信号处理领域中一个前沿的研究分支。
There is long time of speech signal processing and it presents the extensive practical field.
语音信号处理有广泛的应用领域，也有较长的研究历史。
CIS algorithm of cochlear implants' speech signal processing is achieved through hardware method.
文中采用硬件方法对人工耳蜗语音信号处理的CIS算法加以实现。
In this paper, a new linear prediction model in autocorrelation domain for speech signal is presented.
提出了一种在自相关域对语音信号进行线性预测分析的方法。
It presents a new method on reducing additive white noise in speech signal using linear prediction error.
在此提出一种利用线性预测误差去除语音中的加性白噪声的方法。
But it needs prior knowledge of speech signal and its classification and decision making capability is weak.
但它存在需要语音信号的先验知识，分类决策能力弱等缺点。
So, in this chapter, the basic principle and algorithm of LPC for speech signal have been discussed in detail.
所以这一章较为详细地讨论了语音信号线性预测编码的基本原理和算法。
Describes an algorithm for marking pitch in the location of the highest amplitude of continuous speech signal.
阐述了一种连续语音信号的最大幅度的基音标注算法。
Speech signal acquisition and processing is the precondition and foundation of human-computer speech interaction.
语音信号的采集和处理是计算机语音交互的前提和基础。
The speech signal is a analog one, and we can perform its digital transmission by sampling, quantizing and coding.
话音信号是一个模拟信号，通过采样、量化和编码，我们可以实现话音信号的数字传输。
This paper proposed a new method on reducing additive white noise in speech signal using linear prediction residual.
提出了一种利用线性预测残差去除语音中加性白噪声的方法。
We utilize GAs to search stimulative code word in CELP when we have had the knowledge of speech signal coding and GAs.
在具备了语音信号编码和遗传算法的基本知识后，将遗传算法运用于CELP的码字搜索。
One of main services in mobile communication is speech and the key technology is the coding and decoding of speech signal.
语音业务一直是移动通信中的主要业务之一，其中最关键的技术就是语音信号的编码与解码。
Speech enhancement tries to extract clean speech signal from original one with noise and improve the SNR of speech signal.
语音增强则从含噪信号中提取干净的语音信号，提高语音信号的信噪比。
Speech signal, carrying a large amount of information, is so complex with the character of non-stationary and time-varying.
携带着大量信息的语音信号本身是非常复杂的，并且具有非平稳性、时变性等特征。
This paper presents a new formulation for recursively computing the reflection (Partial Correlation) coefficients of speech signal.
本文提出一种计算反射系数的高效公式——位内递推算法。
Based on NLMS, this paper proposes a decorrelation NLMS algorithm (DC-NLMS) by employing auto-correlation of the input speech signal.
在NLMS算法的基础上，从语音信号相关性的角度出发，提出了一种去相关NLMS算法(DC -NLMS)。
Speech signal keeps stability in a short time, and this "short - time characteristic" is an important characteristic of speech signal.
在一短段时间里，语音信号保持相对稳定，这种“短时性”是语音信号的重要特性。
The function of the speech processor lies in dealing with speech signal and producing the pulse which electrodes need to stimulate cochlea.
语音处理器主要作用在于声音的采集处理和产生电极所需的电刺激信号。
To hide information in speech signal, which is a special kind of audio signals, one must consider the characteristics of human audio system.
语音信号作为一种特殊的信息隐藏载体，在其中进行信息隐藏时必须充分考虑人类的听觉特性。
The proposed method uses the maximum singular value of an cumulant matrix to distinguish between voiced parts of the speech signal and noise.
提出一种采用累积量矩阵的最大奇异值来实现语音端点检测的方法，并引入一种自适应的实现方法。
Speech signal is a major channel for human to communicate and to transmit information, making machines understand human language is human's dream.
语音信号是人类交流传递信息的主要途径，让机器听懂人类的语言是人类的梦想。
People are no longer satisfied with the simple speech signal and transmission of the characters data, but expect use many kinds of expression media.
人们不再满足于单纯的语音信号和文字数据的传输，而是期盼使用多种表示媒体。
Speech signal spectrum analysis is the time-varied spectrum analysis. FFT technique on speech signal spectrum analysis are discussed in detail in this paper.
针对语音信号频谱分析实际上是时变频谱分析的特性，详细地讨论了用FFT技术对语音进行频谱分析过程中的方法问题。
Speaker Recognition is one of current researching hotspots of speech signal processing, which has many advantaged superiority at identity authentication field.
说话人识别是当前语音信号处理的研究热点之一，它应用在身份认证领域具有得天独厚的优势。
Since the widely used Hidden Markov model (HMM) in speech recognition is first order Markov model, it can not fully model the temporal dependence of speech signal.
由于在语音识别中被广泛应用的隐马尔可夫模型(HMM)是一重马尔可夫模型，它不能充分地描述语音信号的时间相依性。
Based on the statistical experiments, this paper proposes that the feature variables which indicate speech signal are not always strictly obey normal distribution.
本文在统计实验的基础上指出，表示语音信号的特征变量往往并不严格服从正态分布。
This filter can effectively restrain background noise and acquire clear speech signal when applied in speech signal pick-up from the background with complex noise.
将此滤波器应用于复杂噪音背景的话音信号提取，能很好地抑制背景噪声，从而获得清晰的话音信号。
In this paper, we use full pole model to obtain speech signal LPC, then deduce it's LPCC, and we use the LPCC difference to describe speaker's track dynamic movement.
本文应用全极点模型，提取语音信号的线性预测系数，并推导出其倒谱系数，获得线性预测倒谱差分，用以描述说话人声道的动态变化。
In this paper, we use full pole model to obtain speech signal LPC, then deduce it's LPCC, and we use the LPCC difference to describe speaker's track dynamic movement.
本文应用全极点模型，提取语音信号的线性预测系数，并推导出其倒谱系数，获得线性预测倒谱差分，用以描述说话人声道的动态变化。
应用推荐