语谱图共振峰的自动跟踪算法

本文介绍一个共振峰轨迹的自动跟踪算法，其特点是不借助于其它的信息来源，仅仅是基于语谱图信息，来确定语谱图上前四个共振峰频率的位置和它们关于时间轴的轨迹。算法由三个层面构成；第一层面是进行频率分布的分析，以决定一个最佳的共振峰搜索起始位置；第二层面是采用双搜索算法，来跟踪随时间变化的共振峰轨迹；第三层面是解决某些冲突现象，在共振峰跟踪问题上的难点之一是处理多个共振峰的合并现象和冗余峰值现象。     

石墨点阵柱状光子晶体共振腔的温度特性

构造了一种二维GaAs石墨点阵柱状光子晶体共振腔. 利用时域有限差分方法计算了这种光子晶体共振腔TE_y模的共振峰随温度变化的情况. 计算结果表明这种结构的光子晶体共振腔的共振峰主峰随温度呈分段线性变化趋势,且具有1.60 nm/℃以上的温度响应灵敏度. 同时,计算结果显示,这种结构的光子晶体共振腔具有很好的频率开关特性. 最后,利用计算结果解释了频率开关特性的成因. 关键词： 光子晶体 温度 共振腔 开关     

吸收峰混叠的太赫兹光谱区间组合特征提取算法

太赫兹光谱是物质识别的前沿方法之一。由于不同物质的分子组成或结构各异,许多物质的太赫兹吸收谱会在特定频率上出现吸收峰,可以作为混合物成分检测的重要特征。有效准确地提取这些吸收峰的参数,是提高识别率的关键。多峰拟合算法将光谱曲线拟合成若干个标准峰函数之和,能够同时提取到吸收峰的频率、峰高、峰宽等信息。但是该算法以寻峰算法结果为基础确定吸收峰的大致位置和数量,寻峰结果不一定是最优的拟合结果,而且很难准确识别定位混叠状态的吸收峰。为了提高混叠光谱中吸收峰的识别定位精度,提出以大幅度平滑后的曲线波谷为分界点,将预处理后的光谱分成若干个子区间。然后将子区间组合起来进行多峰拟合,通过遗传算法得到最优的拟合子区间组合和吸收峰频率近似值,拟合时每个子区间中通过峰数递增最优化方法确定拟合的吸收峰数,最后微调优化得到最优的吸收峰频率、峰高值。为了实现物质的识别,通过密度聚类算法得到同一类纯净物在多次测量中的共同吸收峰,以此作为标准数据,通过提出的基于吸收峰特征的光谱匹配算法实现了纯净物和不同含量混合物的快速识别。对10类纯净物的实际光谱数据进行拟合聚类,得到其吸收峰参数,结果与太赫兹光谱数据库一致。通过识别算法对纯净物测试集进行识别的识别率为100%,证明了特征提取和物质识别算法的有效性。对于含有混叠峰的混合物光谱,二阶导数法对葡萄糖-乳糖混合物光谱中被掩盖吸收峰(1.280 THz)的识别率仅为70%,提取到的频率平均值为1.316 THz;而该算法提高识别率至95%,频率平均值为1.281 THz,该算法提高了对混叠峰的分辨能力,能够精确定位混叠峰。对10类纯净物构成的6类不同程度混叠的二元混合物前二、三识别率分别达到90.8%和98.3%,提取到的特征能够有效应用于混合物的成分检测。该算法能够以纯净物数据为标准数据实现成分各异的混合物成分检测,对于太赫兹光谱混合物成分检测有重要意义。     

一类五次方振子系统的叉形分叉及振动共振研究

研究了一类具有分数阶导数阻尼的五次方振子系统中的叉形分叉及振动共振现象. 基于快慢变量分离法, 消去系统中的高频激励成分, 得到关于慢变量的等效系统, 根据等效系统中稳态平衡点的变化情况研究了系统的叉形分叉现象. 结果表明: 高频信号幅值的递增变化会引起亚临界叉形分叉, 高频信号频率和分数阶导数阻尼阶数的递增变化都会引起超临界叉形分叉; 振动共振和叉形分叉是关联的, 当叉形分叉发生时, 振动共振曲线会出现两个峰值, 否则只会出现一个峰值. 通过解析结果和数值模拟结果的对比, 验证了解析分析的正确性. 关键词： 亚临界叉形分叉 超临界叉形分叉 分数阶导数阻尼 振动共振     

基于辐射度的光谱校正与自适应拟合相结合的SPR共振波长确定方法

表面等离子体共振(SPR)传感技术因其灵敏度高、无需标记、可原位实时监测的特点,被广泛应用在生物医学、化学检测、食品安全、环境监测等领域。对于波长调制型SPR传感器,由于光谱测量系统各部件对不同波长响应度不同,从而对光谱共振波长的实际位置产生干扰,影响到实验检测结果。为消除这种影响,提出了一种基于辐射度空间的SPR相对反射率校正方法,并使用阶数自适应拟合算法对共振波长进行精确测定。通过标定光谱测量系统的仪器响应曲线,将光谱仪得到的原始光谱转换成辐射光谱,进而计算出不依赖于光谱仪系统中任何部件的SPR相对反射率。相比于传统的反射率校正方式(以共振光谱除以非共振光谱的结果作为反射率光谱),此方法校正后的光谱在光谱形状上具备半峰宽更窄和共振峰更对称的优势,从而为精确确定共振波长提供了可靠的保障。随后基于拟合误差最小化的目标,对共振区域采用阶数自适应多项式拟合算法确定共振波长。实验测量了不同入射角度下的SPR共振光谱,使用此方法校正得到反射率光谱的半峰全宽保持在(100±10) nm,表明此方法的光谱峰形优势具备普适性;连续采集了4 000组光谱并计算共振波长,其相对标准偏差为0.007 8%,处理速度为每个光谱12 ms,表明此方法具备良好的抗噪声波动鲁棒性和光谱数据处理的实时性;测量了不同浓度下NaCl溶液的SPR共振光谱,其共振波长与溶液折射率线性相关系数为0.998 5,品质因数为传统校正方式的3倍,表明此方法具备良好的可靠性,并且从光谱的处理算法层面提高了传感性能,相比于从工艺制备上提高品质因数的传统方式,处理简便,效果更突出。结果表明,基于辐射度空间的相对反射率矫正与阶数自适应拟合相结合的SPR共振波长检测方法具备可靠性好、运算速度快、分辨率高、抗噪声、品质因数高等特点,能够有效提高SPR传感器的数据处理与传感性能。     

管束穿孔板的管腔耦合共振吸声机理研究

为了揭示管束穿孔板共振吸声结构的吸声机理,利用热黏性条件下基于有限元算法的管束穿孔板仿真模型,研究了平面声波正入射条件下,管束穿孔板内部声场分布特征,并利用阻抗管对吸声系数的理论仿真结果进行了试验验证.结果表明,管束穿孔板在低频主要靠腔体共振吸声,在高频主要靠管共振吸声,管束穿孔板整体呈现出较为明显的管腔耦合共振吸声特征。管束穿孔板共振时管中声强和质点法向振速较大,高频次吸声峰频点处管中和腔中均有驻波形成,频率越高驻波数量越多.管束穿孔板的耦合共振受到管长、腔深、穿孔率和管内径等参数变化的影响,管长对高频耦合共振的影响最大,管长增大使高频主吸声峰频点移向低频,并使相邻主吸声峰之间的间距减小.      

基于高斯拟合的光纤型SPR信号的峰值检测算法

光纤型表面等离子共振光谱具有半峰宽较宽,共振峰不尖锐等特征,传统的针对棱镜型SPR光谱的峰值检测算法无法准确定位此类光谱的共振峰值。为了准确计算光纤型SPR光谱的共振波长,提出了一种基于信赖域算法的高斯拟合方法,在拟合的光谱曲线上利用一维搜索的方式确定峰值位置。通过对标准甘油光谱数据的处理,证明高斯拟合法能够适应光纤型SPR光谱的特征,波长计算准确。通过搭建的SPR系统,实验测量了不同浓度下蔗糖溶液的SPR光谱数据,并分别采用加权质心法,追踪质心法和该研究提出的高斯拟合法进行共振波长的计算。结果表明高斯拟合法能够有效提升分辨率,且运算速度较快,有利于工程集成。     

基于二阶导数谱与特征吸收窗的红外光谱定量检测方法

红外光谱分析技术已被液体、固体和气体混合物各组分的定量检测广泛应用。针对红外光谱定量检测过程中存在的计算过程复杂、受背景及噪声干扰、多组分谱线混叠的问题,一方面,通过获取谱线二阶导数去除背景及噪声干扰,并一定程度上区分混叠谱峰。在获取二阶导数谱线过程中运用Savitzky-Golay滤波器进行平滑滤波处理,并根据谱线频率特征选取Savitzky-Golay滤波器最佳滤波参数,解决了滤波参数缺乏标准化选取方法的问题。另一方面,根据被测组分的吸收分布情况生成特征吸收窗分别对待测原始吸收谱线以及二阶导数谱线加以处理,提取出在浓度计算过程中占据重要性更大的特征吸收的区域,从而排除非吸收区域中背景、噪声及其他组分对计算结果的干扰。以一组丙烷、丙烯和甲苯混合气体为例,运用二阶导数方法以及特征吸收窗方法进行定量检测,并与直接利用原始谱线的检测结果进行比较分析。实验结果表明采用二阶导数谱与特征吸收窗的定量检测方法精度更高,相对误差基本在5%以内。     

过阻尼谐振子的随机共振

研究了由交叉相关高斯白噪声驱动的过阻尼谐振子的随机共振,其中加法噪声被周期信号所调制,运用平稳关联函数的傅里叶变换,导出了过阻尼谐振子随机模型信噪比的精确表达式.结果揭示:在过阻尼谐振子的随机模型中存在二类随机共振.一类随机共振表现为信噪比随乘法噪声强度Q变化的曲线存在共振峰,另一类随机共振表现为信噪比随振子频率ω变化的曲线存在共振峰.大幅度改变信号频率Ω值的大小,信噪比随乘法噪声强度Q变化的曲线有单峰,一峰一谷和单调变化三种不同的形式.     

基于近似熵测度的自适应随机共振研究

变步长随机共振算法有效解决了绝热近似大参数条件下的弱信号检测问题.基于信号近似熵测度的自适应随机共振，实现了变步长随机共振最优输出的自适应求解.周期信号的近似熵不受其幅值和相位变化的影响，而只与其频率及信噪比有关.因此，按照原始数据的采样条件，构造待检测频率在预定信噪比下的标准信号，并以其近似熵为基准，通过自动调节非线性系统的结构参数和计算步长，求得系统输出的近似熵距离矩阵.该矩阵中的最小值所对应的即为自适应条件下非线性动力系统的最优参数.     

Study of the formant and duration in Chinese whispered vowel speech

Study on the acoustical characteristic is important to speech and speaker recognition in Chinese whispered speech. In this paper, the characteristics of whispered speech are introduced and the acoustical characteristics in Chinese whispered speech are discussed. There is no fundamental frequency in the whispered speech, so other characteristics such as the duration and frequency of formant are extracted and analyzed. From experiments with six simple Chinese whispered vowels, it is proved that the duration and the frequency of formant can be used as the main acoustical characteristics in the Chinese whispered recognition.     

Acoustic analysis of the vowel system in Hotan dialect and its contribution to dialect recognition of Uyghur dialects

Based on the actual needs of speech application research such as speech recognition and voiceprint recognition,the acoustic characteristics and recognition of Hotan dialect were studied for the first time.Firstly,the Hetian dialect voice was selected for artificial multi-level annotation,and the formant,duration and intensity of the vowel were analyzed to describe statistically the main pattern of Hetian dialect and the pronunciation characteristics of male and female.Then using the analysis of variance and nonparametric analysis to test the formant samples of the three dialects of Uygur language,the results show that there are significant differences in the formant distribution patterns of male vowels,female vowels and whole vowels in the three dialects.Finally,the GUM-UBM(Gaussian Mixture Model-Universal Background Model),DNN-UBM(Deep Neural Networks-Universal Background Model) and LSTM-UBM(Long Short Term Memory Network-Universal Background Model) Uyghur dialect recognition models are constructed respectively.Based on the Mel-frequency cepstrum coefficient and its combination with the formant frequency for the input feature extraction,the contrastive experiment of dialect i-vector distinctiveness is carried out.The experimental results show that the combined features of the formant coefficients can increase the recognition of the dialect,and the LSTM-UBM model can extract more discriminative dialects than the GMM-UBM and DNN-UBM.     

Long-Term-Average Spectrum Characteristics of Country Singers During Speaking and Singing

Five premier male country singers involved in our previous studies spoke and sang the words of both the national anthem and a country song of their choice. Long-term-average spectra were made of the spoken and sung material of each singer. The spectral characteristics of county singers' speech and singing were similar. A prominent peak in the upper part of the spectrum, previously described as the "speaker's formant," was found in the county singers' speech and singing. The singer's formant, a strong spectral peak near 2.8 kHz, an important part of the spectrum of classically trained singers, was not found in the spectra of the country singers. The results support the conclusion that the resonance characteristics in speech and singing are similar in country singing and that county singing is not characterized by a singer's formant.     

The acoustic bases for gender identification from children's voices.

The purpose of this study was to examine the acoustic characteristics of children's speech and voices that account for listeners' ability to identify gender. In Experiment I, vocal recordings and gross physical measurements of 4-, 8-, 12-, and 16-year olds were taken (10 girls and 10 boys per age group). The speech sample consisted of seven nondiphthongal vowels of American English (/ae/ "had," /E/ "head," /i/ "heed," /I/ "hid," /a/ "hod," /inverted v/ "hud," and /u/ "who'd") produced in the carrier phrase, "Say /hVd/ again." Fundamental frequency (f0) and formant frequencies (F1, F2, F3) were measured from these syllables. In Experiment II, 20 adults rated the syllables produced by the children in Experiment I based on a six-point gender rating scale. The results from these experiments indicate (1) vowel formant frequencies differentiate gender for children as young as four years of age, while formant frequencies and f0 differentiate gender after 12 years of age, (2) the relationship between gross measures of physical size and vocal characteristics is apparent for at least 12- and 16-year olds, and (3) listeners can identify gender from the speech and voice of children as young as four years of age, and with respect to young children, listeners appear to base their gender ratings on vowel formant frequencies. The findings are discussed in relation to the development of gender identity and its perceptual representation in speech and voice.     

Using the method of modified linear prediction for robust acoustic coding of speech

Ensuring the acceptable acoustic coding of a signal under the conditions of noise interference is an important problem of recognizing speech signals. The result of calculating acoustic characteristics by the existing algorithms strongly depends on the signal-to-noise ratio. To solve the problem related to ensuring robust acoustic coding, we propose an algorithm for obtaining cepstral coefficients on the basis of the method of modified linear prediction. The algorithm is aimed at the specific features of the vocalized speech signals with a pronounced formant structure including the case of wideband additive interference. The algorithm stability as a function of the noise level is studied experimentally. The result is improved compared with the well-known approaches. __________ Translated from Izvestiya Vysshikh Uchebnykh Zavedenii, Radiofizika, Vol. 49, No. 7, pp. 593–598, July 2006.     

提升小波加权自相关函数的基音检测算法*

随着计算机技术的发展,语音信号处理作为人机交互的重要渠道,其在复杂噪声环境下的特征值检测算法直接关系到计算机的运算效率。基音周期是语音特征值提取的重要参数之一。针对传统基音检测算法在噪声环境下检测精度低的问题,提出了一种基于自适应提升小波变换加权线性预测误差自相关函数的基音检测算法。该方法用多级提升小波近似系数加权求和的方法来弥补自相关函数随着时间延迟量的增加幅值衰减的缺陷;用线性预测误差自相关函数的方法来抑制共振峰的干扰,然后将两种方法结合来突出基音周期处的峰值。实验结果表明,与传统的自相关函数法和小波加权法相比,该方法能有效减弱共振峰的影响,突出基音周期处的峰值,提高基音周期检测精度,鲁棒性更好。     

Formant trajectory characteristics of males with amyotrophic lateral sclerosis.

The purpose of this study was to describe the formant trajectories produced by males with amyotrophic lateral sclerosis (ALS), a degenerative neuromuscular disease that is typically associated with dysarthria. Formant trajectories of 25 males with ALS and 15 neurologically normal geriatric males were compared for 12 words selected from the speech intelligibility task developed by Kent et al. [J. Speech. Hear. Disord. 54, 482-499 (1989)]. The results indicated that speakers with ALS (1) produced formant transitions having shallower slopes than transitions of normal speakers, (2) tended to produce exaggerations of formant trajectories at the onset of vocalic nuclei, and (3) had greater interspeaker variability of formant transition characteristics than normal speakers. Within the group of ALS speakers, those subjects who were less than 70% intelligible produced distinctly more aberrant trajectory characteristics than subjects who were more than 70% intelligible. ALS subjects who were less than 70% intelligible produced many trajectories that were essentially flat, or that had very shallow slopes. These results are discussed in terms of the speech production deficit in the dysarthria associated with ALS, and with respect to the potential influence of aberrant trajectories on speech intelligibility.     

Time-varying spectral change in the vowels of children and adults

Recent studies have shown that time-varying changes in formant pattern contribute to the phonetic specification of vowels. This variation could be especially important in children's vowels, because children have higher fundamental frequencies (f0's) than adults, and formant-frequency estimation is generally less reliable when f0 is high. To investigate the contribution of time-varying changes in formant pattern to the identification of children's vowels, three experiments were carried out with natural and synthesized versions of 12 American English vowels spoken by children (ages 7, 5, and 3 years) as well as adult males and females. Experiment 1 showed that (i) vowels generated with a cascade formant synthesizer (with hand-tracked formants) were less accurately identified than natural versions; and (ii) vowels synthesized with steady-state formant frequencies were harder to identify than those which preserved the natural variation in formant pattern over time. The decline in intelligibility was similar across talker groups, and there was no evidence that formant movement plays a greater role in children's vowels compared to adults. Experiment 2 replicated these findings using a semi-automatic formant-tracking algorithm. Experiment 3 showed that the effects of formant movement were the same for vowels synthesized with noise excitation (as in whispered speech) and pulsed excitation (as in voiced speech), although, on average, the whispered vowels were less accurately identified than their voiced counterparts. Taken together, the results indicate that the cues provided by changes in the formant frequencies over time contribute materially to the intelligibility of vowels produced by children and adults, but these time-varying formant frequency cues do not interact with properties of the voicing source.     

Formant frequencies, bandwidths, and Qs in helium speech

We used linear prediction analysis to estimate a diver's vocal tract response for isolated vowels spoken in air at 1 atm and in heliox at simulated depths of 54-, 120-, 300-, and 500-m seawater. We specifically measured formant frequency, bandwidth, and Q changes as a function of environment. The formant frequencies shifted upward nonlinearly in helium speech. The formant bandwidths in helium speech increased as much as 14 times their corresponding bandwidths in normal speech. The ratios of formant Qs (helium speech to normal speech) ranges from 0.3 (low formants) to 2 (high formants).