流形学习在运动声源声特征提取方面的研究*

运动声源因声信号时变性、叠加性和空时耦合性强,声数据呈现高维、非线性等特点,使得关键声特征提取困难,声特征提取方法复杂度高、数值计算量大、有效性差。因此,如何有效提取声特征并降低提取方法复杂度成为目前多源声场声源精准识别需迫切解决的关键科学问题。由此,该文提出短时傅里叶变换(STFT)和局部线性嵌入算法(LLE)联合的STFT-LLE流形学习声特征提取方法,并将此方法应用于运动声特征提取,且通过仿真实验测试对其进行了验证。该方法为运动声目标的分类识别提供了技术支撑。     

Cell image recognition with radial harmonic Fourier moments

Various types of moments have been utilized to recognize image patterns in a number of applications. Multi-distortion invariant－radial harmonic Fourier moments were investigated in the recognition of cell smear images. Image recognition experiments showed that radial harmonic Fourier moments are not only highly concentrated image features but also have a good property of antidistortion and antinoise. They may be used in fast and accurate automatic cell recognition.     

加筋板材料辨识的导纳特征表达与冲击声特征提取

以表征物理属性的导纳特征为中间量,提取与加筋板材料属性有关的冲击声特征。先用相关分析方法获得金属加筋板物理属性的导纳特征表达以及导纳特征与冲击声特征之间的联系,间接得到表征声源物理属性的冲击声特征,然后通过支持向量机分类器验证不同特征在金属加筋板材料分类辨识中的性能。结果表明,所得的4组冲击声特征能准确识别出不同的材料,单个特征的识别率与对应材料属性的可分程度有关,理想冲击声声特征比音色特征的平均识别率更高。由此可见,利用导纳特征提取与材料属性相关冲击声特征的方法是有效的,且所提的特征能够很好的反映声源材料属性。     

利用白光干涉技术测量块状材料的群折射率

对群折射率的精确而快速测量是光学工程领域一个基础而又亟需解决的难题,本文提出了一套光谱型迈克尔逊白光干涉系统的解决方案.该系统充分利用了微型光纤光谱仪一次测量便可获得所有干涉信息的特点,无需机械扫描装置,具有结构简单和测量快速的优点.与传统的窗口傅里叶变换算法相比,本文采用小波变换直接从干涉信号的小波脊中提取群延迟,减小了由相位求导得到群延迟过程中引入的误差放大效应,进而提高了群折射率的测量精度.基于此迈克尔逊白光干涉系统,在不同干涉位置处对两块不同厚度的石英和BK7玻璃进行了测量,实验结果表明此方法在宽 关键词： 白光干涉 群折射率 小波变换 傅里叶变换     

冲击声的稀疏特征提取及声源类型识别

针对低信噪比下声源材料类型的细分任务,将稀疏表达用于冲击声信号的声源类型识别,提取的稀疏特征相比传统的MFCC特征有效改善了识别性能。分别基于3种预定义词典和一组根据训练信号学习的词典,利用正交匹配追踪(OMP)方法对录制冲击声进行稀疏表达,提取稀疏特征用于不同信噪比下冲击声信号的声源辨识,并与MFCC特征进行比较。对包含12类材料的冲击声数据库的分类结果显示,在几乎所有情况下,稀疏特征比MFCC特征具有更好的识别效果。特别是在信噪比较低的情况下,稀疏特征具有更好的抗噪性能。     

一种用于光纤链路振动信号模式识别的规整化复合特征提取方法

相位光时域反射链路监测系统是一种利用光纤作为传感介质的传感系统, 能够监测、定位、识别入侵信号.模式识别模块是其重要组成部分, 实时智能区分安全扰动和危险入侵.本文提出一种用于光纤链路振动信号模式识别的复合特征提取方法.利用改进的双门限方法确定有效信号段的起止位置, 结合最大能量与最高信噪比挑选出采样周期内主要入侵扰动的特征段.综合利用特征段时域持续时间和小波包能量谱提取复合特征向量, 使用支持向量机进行模式识别.实验表明, 基于本文提出的规整化特征提取方法的模式识别准确率有了显著提高.     

基于流形学习的声目标特征提取方法研究

为解决目前声目标识别面临的鲁棒性不足问题,提出将流形学习应用到声目标的特征提取中,在经典流形学习算法的基础上,研究讨论了目标声信号频域中存在的低维流形,通过两种实际的地面和低空飞行声目标数据集进行对比识别实验,分析了基于流形学习的声目标特征提取方法的性能,结果表明基于流形学习的特征提取方法可以发现声信号的本质特征,提高了声目标识别系统的准确性和鲁棒性. 关键词： 声目标识别 特征提取 流形学习     

基于小波包变换的复合材料超声波检测信号特征提取

本文以碳纤维复合材料常见缺陷分层、孔隙、疏松的超声波检测缺陷信号为研究对象，对超声波检测信号进行小波包变换，提取包含信号绝大部分能量的近似系数波形特征及细节系数的统计量作为样本的特征值。应用BP神经网络分类器进行分类识别验证，取得较好的识别效果。该方法能以较小的特征维数表征原始信号特点。     

窗函数在扬声器异常声客观检测中的影响

该文研究了基于短时傅里叶变换的扬声器异常声检测方法中,窗函数对特征距离阈值设定的影响。以连续对数扫频信号激励被测扬声器单元,并对响应信号进行短时傅里叶变换。通过对比不同窗函数下,被测单元之间的特征距离来寻找合适的窗函数。结果表明,不同窗函数下得到的特征距离具有不同的离散度;选择合适的窗函数,可使特征距离具有较大离散度,从而更合理地设定阈值。     

基于Hilbert-Huang变换的水声信号特征提取及分类技术

水下目标噪声信号是一种典型的非线性非平稳随机信号，Hilbert-Huang变换较传统方法在处理此类信号时具有很大优势，据此本文提出了一种基于Hilbert-Huang变换的水下目标噪声特征提取方法。本文将环境噪声当作一类目标，利用上述方法对涉及两种航行船舶、海洋生物和海洋环境噪声的四类目标信号的特征进行了详细分析，最后对各类目标噪声信号进行了分类识别实验。实验结果表明基于Hilbert-Huang变换提取的特征对各类目标具有较好的可分性，达到了较高的识别率。     

Feature extraction for sound source material recognition with impact sounds

Sound source recognition is a part of environmental sound recognition,which is one of the most important research areas in pattern recognition.Impact sounds carry much physical information associated with the sound sources,which makes impact sound based sound source recognition an important approach to improve recognition performance.In this study,the impact sound continuum synthesized with a ball-plate collision model is used for material recognition of the impacted plates.The basis function learning method and time-frequency representation methods,including the short time Fourier transform and the wavelet packet transform,are applied into classification and the recognition results are compared.The result shows that the features obtained by using the basis function learning perform better for material classification of the impacted plates than that by using the short time Fourier transform and the wavelet packet transform.This demonstrates the high efficiency and superiority of this method in material recognition of sound sources.     

Individual differences in source identification from synthesized impact sounds

Impact sounds were synthesized according to standard textbook equations given for the motion of freely vibrating membranes, bars, and plates. In a two-interval, forced-choice procedure, highly practiced listeners identified from these sounds predefined target sources based on their material and size, the hardness of the striking mallet, and the presence or absence of light damping applied to the center of the source. Listener decision strategy in each case was determined from a discriminant analysis of trial-by-trial responses resulting in a vector of regression weights given to different acoustic parameters. The analysis revealed significant differences in decision strategy across listeners within identification task, but similarity in decision strategy within listeners across variations in task. Only when the acoustic information for identification was highly constrained (identification of damping) did listeners adopt similar decision strategies approaching that of an ideal observer. Despite the large individual differences in decision strategy, identification accuracy was, in most cases, similar across listeners. Where there were differences in identification accuracy the differences appeared largely related to differences in internal noise and not decision strategy. The results are generally comparable to those obtained for the discrimination of arbitrary tone patterns.     

基于平板冲击声的声源特性表征及自动识别*

以平板结构导纳函数为纽带,建立冲击声信号特征与声源特性之间的关联,获得与声源属性密切相关的特征用于目标分类。针对四边简支矩形被击板,借助信号参数识别算法获得与声源物理属性有关的6维导纳特征,并从冲击声样本中提取80维音色特征,将音色特征和导纳特征做相关性分析,获得与声源物理属性相关的信号特征集。利用BP神经网络进行分类,结果表明,当采用与特定声源物理属性相关的信号特征子集时,分类效果达到同组最优。     

The admittance feature representation and impact sound feature extraction in the material identification of ribbed plates

The admittance features representing the physical attributes are used as the intermediates to extract the material-attributes-related impact sound features of ribbed plates.Firstly, the admittance feature representations of metal ribbed plates attributes are obtained and the relationship between the admittance features and the impact sound features are established via correlation analysis method. Then, material-attributes-related impact sound features are obtained indirectly. Finally, the performances of different sound features for the material recognition of ribbed-metal plates are verified through the Support Vector Machine classifier.The results indicate that the obtained four sets of features can effectively identify the materials of the metal ribbed plates, while the accuracy of a single feature depends on the separable degree of the corresponding material attribute. And the features extracted based on admittance functions have higher average accuracy than that of timbre features. Therefore, the proposed sound feature extraction method based on admittance features is valid, and the extracted sound features can effectively reflect the physical attributes.     

人工耳蜗的环境声识别研究

对14位正常听力者开展了环境声的人工耳蜗仿真声识别实验,比较了两类声码器仿真(正弦载波和噪声载波)条件下的环境声识别效果差异,然后对9位讲普通话的成年人工耳蜗植入者开展了环境声识别实验。实验材料是从互联网上搜集,并经过12位正常听力者主观测试验证后,筛选出的67种环境声。结果显示,载波类型没有对67种环境声的平均识别效果产生显著影响,但是声学特征的差异会导致单个环境声的识别效果对载波类型有依赖。另外,人工耳蜗植入者的环境声识别效果较差,有待通过信号处理策略、神经接口和康复手段的改进而得到提高。本研究中开发的环境声材料可以用于评估人工耳蜗环境声识别效果。     

声源定位实验——一个具有明确应用背景和前沿技术特点的综合性实验

研制开发了一种进行声源定位的实验装置.利用波的传播特性,可以推知物体的空间位置.文章介绍了声发射平面定位的原理,自制的实验装置和实验结果.     

Sound field of a baffled sound source covered by an anisotropic rigid-porous material

A theoretical approach for the sound field of a piston sound source covered by a finite thickness layer of anisotropic rigid-porous material is presented. The formulation is an extension of the method worked out by Amedin et al. [Sound field of a baffled piston source covered by a porous medium layer. J Acoust Soc Am 1995;98(3):1757]. First, in the present study the sound field of a point source is described by cylindrical waves. Then, with the proper boundary conditions, the sound pressure radiated from a piston source covered by a layer of anisotropic porous material can be calculated. The effects of frequency and bulk density of material on the sound propagation in an anisotropic porous material are studied. Finally, the effect of anisotropy is discussed.