Vowel-Based Non-uniform Prosody Modification for Emotion Conversion

The objective of this work is to develop a rule-based emotion conversion method for a better emotional perception. In this work, performance of emotion conversion using the linear modification model is improved by using vowel-based non-uniform prosody modification. In the present approach, attempts were made to integrate features like position and identity for addressing the non-uniformity in prosody generated due to the emotional state of the speaker. We mainly concentrate on the parameters such as strength, duration and pitch contour of vowels at different parts of the sentence. The influence of emotions on the above parameters is exploited to convert the speech from neutral emotion to the target emotion. Non-uniform prosody modification factors for emotion conversion are based on the position of vowels in the word, and the position of the word in the sentence. This study is carried out by using Indian Institute of Technology-Simulated Emotion speech corpus. Evaluation of the proposed algorithm is carried out by a subjective listening test. From the listening tests, it is observed that the performance of the proposed approach is better than the existing approaches.     

Improving the Flexibility of Dynamic Prosody Modification Using Instants of Significant Excitation

Modification of suprasegmental features such as pitch and duration of original speech by fixed scaling factors is referred to as static prosody modification. In dynamic prosody modification, the prosodic scaling factors (time-varying modification factors) are defined for all the pitch cycles present in the original speech. The present work is focused on improving the naturalness of the prosody modified speech by reducing the generation of piecewise constant segments in the modified pitch contour. The prosody modification is performed by anchoring around the accurate instants of significant excitation estimated from the original speech. The division of longer pitch intervals into many equal intervals over long speech segments introduces step-like discontinuities in the form of piecewise constant segments in the modified pitch contours. The effectiveness of proposed dynamic modification method is initially confirmed from the smooth modified pitch contour plot obtained for finer static prosody scaling factors, waveforms, spectrogram plots and comparison subjective evaluations. Also, the average \(F_0\) jitter computed from the pitch segments of each glottal activity region in the modified speech is proposed as an objective measure for the prosody modification. The naturalness of the prosody modified speech using the proposed method is objectively and subjectively compared with that of the existing zero frequency filtered signal-based dynamic prosody modification. Also, the proposed algorithm effectively preserves the dynamics of the prosodic patterns in singing voices where in the \(F_0\) parameters rapidly and continuously fluctuate within a higher \(F_0\) range.     

A novel algorithm for voice conversion using Canonical Correlation Analysis

A novel algorithm for voice conversion is proposed in this paper. The mapping function of spectral vectors of the source and target speakers is calculated by the Canonical Correlation Analysis （CCA） estimation based on Gaussian mixture models. Since the spectral envelope feature remains a majority of second order statistical information contained in speech after Linear Prediction Coding （LPC） analysis, the CCA method is more suitable for spectral conversion than Minimum Mean Square Error （MMSE） because CCA explicitly considers the variance of each component of the spectral vectors during conversion procedure. Both objective evaluations and subjective listening tests are conducted. The experimental results demonstrate that the proposed scheme can achieve better performance than the previous method which uses MMSE estimation criterion.     

A Multi-level GMM-Based Cross-Lingual Voice Conversion Using Language-Specific Mixture Weights for Polyglot Synthesis

For any given mixed-language text, a multilingual synthesizer synthesizes speech that is intelligible to human listener. However, as speech data are usually collected from native speakers to avoid foreign accent, synthesized speech shows speaker switching at language switching points. To overcome this, the multilingual speech corpus can be converted to a polyglot speech corpus using cross-lingual voice conversion, and a polyglot synthesizer can be developed. Cross-lingual voice conversion is a technique to produce utterances in target speaker’s voice from source speaker’s utterance irrespective of the language and text spoken by the source and the target speakers. Conventional voice conversion technique based on GMM tokenization suffer from degradation in speech quality as the spectrum is oversmoothed due to statistical averaging. The current work focuses on alleviating the oversmoothing effect in GMM-based voice conversion technique, using (source) language-specific mixture weights in a multi-level GMM followed by selective pole focusing in the unvoiced speech segments. The continuity between the frames of the converted speech is ensured by performing fifth-order mean filtering in the cepstral domain. For the current work, cross-lingual voice conversion is performed for four regional Indian languages and a foreign language namely, Tamil, Telugu, Malayalam, Hindi, and Indian English. The performance of the system is evaluated subjectively using ABX listening test for speaker identity and using mean opinion score for quality. Experimental results demonstrate that the proposed method effectively improves the quality and intelligibility mitigating the oversmoothing effect in the voice-converted speech. A hidden Markov model-based polyglot text-to-speech system is also developed, using this converted speech corpus, to further make the system suitable for unrestricted vocabulary.     

一种用于语声转换系统的LPC残差信号生成算法

残差信号中的基音信息对语音的说话人个性特征有着重要的影响.本文首先通过转换后的语音谱包络特征参数(LSP)来预测相应的目标基音周期,再利用预测的目标基音周期来修改源语音的残差信号,从而生成所需要的目标语音残差信号.客观评测和主观听觉测试都表明,本文的残差信号生成算法(PP DCT,Pitch Prediction Discrete Cosine Trans-form)性能要好于以往的残差预测法.     

采用模型自适应的语音转换方法

针对非对称语音库情况下的语音转换,提出了一种有效的基于模型自适应的语音转换方法。首先,通过最大后验概率（Maximum A Posteriori,MAP）方法从背景模型分别自适应训练得到源说话人和目标说话人的模型;然后,通过说话人模型中的均值向量训练得到频谱特征的转换函数;并进一步与传统的INCA转换方法相结合,提出了基于模型自适应的INCA语音转换方法,有效实现了源说话人频谱特征向目标说话人频谱特征的转换。通过客观测试和主观测听实验对提出的方法进行评价,实验结果表明,与INCA语音转换方法相比,本文提出的方法可以取得更低的倒谱失真、更高的语音感知质量和目标倾向度;同时更接近传统基于对称语音库的高斯混合模型（Gaussian Mixture Model,GMM）的语音转换方法的效果。      

High resolution pole-zero analysis of Parkinsonian speech

High resolution analysis of voiced speech signals in Parkinsonian patients using a pitch synchronous pole-zero model is introduced. A modified estimation error is defined leading to an accurate and consistent determination of the excitation instants of the model. An infinite resolution in determining these instants, despite the finite sampling interval, is achieved by mapping the discrete (digitized) problem into a continuous one. The proposed analysis was found to be useful in analyzing Parkinsonian speech where the goal was to detect and quantify the Parkinsonian tremor and rigidity from sustained voiced sounds.     

3GPP AMR—NB与ITU—TG．729A语音编码标准技术的对比研究

介绍了AMR—NB与G．729A2种语音编码标准的特点和算法,并就其线性预测分析、基音搜索、代数码书搜索和增益量化4个方面技术进行了比较。在线性预测分析方面主要对2种算法的加窗、LSP量化与内插的不同进行了陈述;在基音搜索方面对开环和闭环基音搜索上的差异进行了分析,并且对代数码书结构和搜索算法及增益参数量化的差别进行了阐述。最后给出了2种编码标准的语音质量、计算复杂度、空间复杂度等性能测试结果。     

采用超音段韵律特征联合短时频谱的语音转换

传统的语音转换方法往往着重于语音的声道特征和基频的转换,而忽视了其他的超音段韵律特征,这导致转换后的语音目标倾向性不够明显,合成语音自然度不高,不能很好地反应说话人个性化特征。本文在短时谱包络转换的基础上,加入了基频、语速、停顿、重音等多种超音段韵律特征进行转换处理,以提高语音转换性能。其中,采用基频目标模型对基音频率建模,然后运用高斯混合模型（GMM）训练得到转换规则,而语速、停顿、重音则采用基于单高斯统计分析的最大似然估计方法训练得到转换规则。实验结果表明,在加入超音段韵律特征转换之后,系统非常明显地提高了转换语音的目标倾向性和自然度。      

一种汉语语音变换技术

汉语语音变换技术的目的是将汉语语音中源说话人的语音特征转换为目标说话人语音特征。提出的适用于汉语说话人的变换算法分为3个部分:前两部分用高斯混合模型实现了语音的谱包络(线性预测编码)及其激励(残差)的转换;第三部分采用支持向量回归算法实现语音的韵律变换规则建模,结合汉语语音特点利用基音同步叠加算法实现语音的超音段特征调整。与现有的语音变换算法进行比较,算法针对汉语语音超音段发音特点进行韵律调整,有效实现了汉语语音变换并得到高自然度合成语音,是一种有效的汉语语音变换算法。     

Channel and source considerations of a bit-rate reduction technique for a possible wireless communications system's performance enhancement

In wireless commercial and military communications systems, where bandwidth is at a premium, robust low-bit-rate speech coders are essential. They operate at fix bit rates and those bit rates cannot be altered without major modifications in the vocoder design. A novel approach to vocoders, in order to reduce the bit rate required to transmit speech signal, is proposed. While traditional low-bit-rate vocoders code original input speech, the proposed procedure operates on the time-scale modified signal. The proposed method offers any bit rate from 2400 b/s to downwards without modifying the principle vocoder structure, which is the new NATO standard, Stanag 4591, Mixed Excitation Linear Prediction (MELP) vocoder. We consider the application of transmitting MELP-encoded speech over noisy communication channels by applying different modulation techniques, after time-scale compression is applied. Three different time-scale modification algorithms have been evaluated and waveform similarity overlap and add (WSOLA) algorithm has been selected for time-scale modification purposes. Computer simulation results, both source and channel, are presented in terms of objective speech quality metrics and informal subjective listening tests. Design parameters such as codec complexity and delay are also investigated. Simulation results lead to a possible wireless communications system, whose performance might be enhanced by using the spared bits offered by the procedure.     

基于线性预测残差倒谱的多语音基音频率检测

本文提出了一种基于线性预测残差倒谱的多语音基音频率检测算法,该算法首先对混合语音信号进行线性预测分析,进而计算预测信号与原混合信号的残差,并对残差信号做倒谱变换,得到混合语音信号的线性预测残差倒谱;然后在该信号的残差倒谱中,结合图像处理的技术,利用语音信号基音倒频匹配法检测出多语音信号的基音频率;最后在基音标定的过程中,本文算法利用语音信号的连续特性,依据信号基音频率前后差距变化最小原则标记出各基音所属话者。实验结果表明,本文提出的算法在弱回声及无回声的情况下能快速有效地从单声道混合语音信号中检测出多语音基音信息。     

基于正弦加噪声模型的说话人转换方法

提出一种基于正弦加噪声模型的说话人转换方法,着重讨论通过修改音素段内的声学参数实现说话人的转换。通过修改基音频率和共振峰结构,该方法合成的语音有效地模拟了目标说话人的特性。听力测试表明,转换后的语音和目标说话人的语音相似度达到78．8％。与经典的LPC方法的对比实验验证了该法在合成语音质量方面的优越性。     

支持向量回归在声音转换中的应用

声音转换是将源说话人的声音转化成具有目标说话人特征信息的声音的方法。将3种不同的回归方法:多项式回归,线性多变量回归以及支持向量回归分别应用于声音转换。实验分别对5个普通话元音进行转换。主观和客观评估了每种方法的语音转换质量。结果表明,支持向量回归具有更强的学习能力,使转换语音具有更好的目标倾向性。与多项式回归和线性多变量回归相比,支持向量回归既提高了泛化能力又避免了频谱不连续性,使转换语音与目标语音的频谱距离失真分别减少了33.29%和35.24%。     

约束条件下的结构化高斯混合模型及非平行语料语音转换

提出一种约束条件下的结构化高斯混合模型及非平行语料语音转换方法.从源与目标说话人的原始非平行语料中提取出少量相同音节,在结构化高斯混合模型的训练过程中,利用这些相同音节包含的语义信息及声学特征对应关系对K均值聚类中心进行约束,并在（Expectation Maximum,EM）迭代过程中对语音帧属于模型分量的后验概率进行修正,得到基于约束的结构化高斯混合模型（Structured Gaussian Mixture Model with Constraint condition,C-SGMM）.再利用全局声学结构（Acoustic Universal Structure,AUS）原理对源和目标说话人的约束结构化高斯混合模型的高斯分布进行匹配对准,推导出短时谱转换函数.主观和客观评价实验结果表明,使用该方法得到的转换后语音在谱失真,目标倾向性和语音质量等方面均优于传统的结构化模型语音转换方法,转换语音的平均谱失真仅为0.52,说话人正确识别率达到95.25%,目标语音倾向性指标ABX平均为0.82,性能更加接近于基于平行语料的语音转换方法.     

Influence of the position of the analysis frame in LPC pitch-synchronous analysis

The paper deals with the effect of the position of the time interval on the evaluation of voiced sound characteristics in speech analysis using the Linear Prediction Coding (LPC) technique. It is shown that when the analysis frame coincides with a single pitch period (pitch-synchronous analysis) an erroneous alignment (with respect to pitch pulses) of the analysis interval may introduce significant errors in the estimation of formant frequencies and bandwidths, and more generally, of sound spectrum. Synthetic speech is used to investigate the phenomenon and the experimental results are discussed. On the basis of these results, various techniques for reducing the influence of the position of the analysis interval on Linear Prediction parameters are discussed.     

Pitch-Scaled Spectrum Based Excitation Model for HMM-based Speech Synthesis

The speech generated by hidden Markov model (HMM)-based speech synthesis systems (HTS) suffers from a ‘buzzing’ sound, which is due to an over-simplified vocoding technique. This paper proposes a new excitation model that uses a pitch-scaled spectrum for the parametric representation of speech in HTS. A residual signal produced using inverse filtering retains the detailed harmonic structure of speech that is not part of the linear prediction (LP) spectrum. By using pitch-scaled spectrums, we can compensate the LP spectrum using the detailed harmonic structure of the residual signal. This spectrum can be compressed using a periodic excitation parameter so that it can used to train HTS. We define an aperiodic measure as the harmonics-to-noise ratio, and calculate a voicing-cut off frequency to fit the aperiodic measure to a sigmoid function. We combine the LP coefficient, pitch-scaled spectrum, and sigmoid function to create a new parametric representation of speech. Listening tests were carried out to evaluate the effectiveness of the proposed technique. This vocoder received a mean opinion score of 4.0 in analysis-synthesis experiments, before dimensionality reduction. By integrating this vocoder into HTS, we improved the sound of the synthesized speech compared with the pulse train excitation model, and demonstrated an even better result than STRAIGHT-HTS.     

基于细粒度韵律建模和条件CycleGAN的非平行蒙古语语音转换方法

语音转换技术在保持语义内容不变的前提下将源说话人的语音音色转换为目标说话人。目前,蒙古语语音转换面临语料匮乏、蒙古语字词在发音上韵律变化丰富等问题。针对这些问题,本文提出一种基于细粒度韵律建模和条件CycleGAN的非平行蒙古语语音转换方法。该方法首先使用连续小波变换提取细粒度的语音韵律特征,然后向CycleGAN中加入说话人向量构建条件CycleGAN,最后使用条件CycleGAN得到源说话人和目标说话人之间稳定的韵律转换。实验结果表明,该方法与传统CycleGAN语音转换方法相比能够有效提升蒙古语语音转换效果,在语音自然度和说话人相似度的MOS评分上分别提升了0.1和0.2。      

语声转换技术发展及展望

语声转换通过改变语音信号的声学特征参数来调整语音的个性特征，从而使得转换后的源说话人语音听起来就像是目标说话人的声音一样。系统地介绍了当前语声转换技术的发展状况，在描述语声转换技术的应用场景和系统框架的基础上，着重阐述了系统的转换模块，即声道特性的转换和韵律转换，特别是重点介绍了声道特性的转换算法。简要地介绍了系统性能的测试方法，最后对全文进行了总结，并针对当前语声转换技术还存在的一些问题，对未来的发展进行了展望。     

声音转换技术的研究与进展

声音转换是一项改变说话人声音特征的技术,可以将一人的语音模式转换为与其特性不同的另一人语音模式.声音转换算法的目标是确定一个什么样的模式转换规则,使转换语音保持第一个说话人原有语音信息内容不变,而具有第二个说话人的声音特点.本文介绍了当前声音转换技术领域的研究状态,主要分析现有声音转换技术中各种转换算法的实现原理,描述声音转换系统性能的各种评估方法,最后给出了对声音转换技术的简要评述和展望.