NRZ码全光帧头识别器研究

提出了一种基于太赫兹非对称光解复用器结构的NRZ码全光帧头识别方案.通过注入辅助的连续光,解决了信号在SOA中的竞争,实现三态到两态的转换.对40 Gb/s信号的数值分析表明,这种方案可获得消光比为13.29 dB的判决识别信号输出.对非线性相移差偏离的影响进行了分析,结果表明如果限制非线性相移差的偏离小于0.1 rad,仍然可获得消光比13 dB以上的判决识别信号输出.对2.5 Gb/s帧头信号的判决实验表明,理论分析结果与实验相符.     

非线性荧光光谱的神经网络分析及其应用

大功率超快脉冲激光和气体相互作用可产生非线性荧光光谱,不同的气体分子具有不同的非线性荧光光谱。因而这种光谱可以作为物质的指纹模式加以识别分类,进而获知气体的成分。由于不同气体分子的光谱在同一波段上有很大的交叉重叠,用传统的光谱分析方法分析存在困难,采用神经网络方法分析上述非线性荧光光谱,利用经过预处理的荧光光谱数据作为模式样本,其中一部分样本作为学习样本对级联神经网络进行训练,用训练好的网络对所有样本进行实时识别,学习样本和测试样本的的正确识别率均可达100％,结果表明此方法可实时判断混合气体的组分。     

远程采集模糊无人机图像的特征识别方法研究

地面对空中无人机的视觉识别中,由于无人机的飞行速度、角度呈现非线性变化。使得采集的疑似图像存在特征模糊、衰退等问题,传统的模式识别方法无法提取无人机图像的主要特征,极大程度上降低了图像的识别概率。提出一种引入球面谐波基图像特征细分的无人机识别算法,建立球面谐波基图像识别模型,利用无人机图像的球面谐波基图像近似率,对模糊图像的差异特征进行依次识别。实验结果表明,利用改进算法建立的模糊无人机图像差异特征识别模型,具有一定的优越性,提高了无人机识别的准确率。     

基于核独立成分分析的人脸识别

研究一种基于核独立成分分析的人脸识别方法。利用支持向量机的核函数思想,将原始人脸图像向量映射到高维特征空间,然后在高维特征空间中进行独立成分分析(ICA),提取非线性独立成分作为特征向量进行分类识别。实验结果表明该方法要比常规的基于ICA和PCA的人脸识别算法的识别率要高。     

分数双重非线性相关器特性研究

将分数傅里叶变换引入到双重非线性相关方法中,通过对参考图像和目标图像分数傅里叶变换谱的双重非线性操作实现非线性分数相关.该方法利用非线性参数以及分数阶控制目标的形状和纹理的权重,从而实现目标相关识别的调控.该非线性分数相关系统由一个光电混合装置实现,数值仿真表明,这种光学图像识别系统的优点在于对目标的形状失真和纹理改变...     

基于神经网络的光谱识别系统的设计与分析

光谱识别技术是光谱定性分析的基础。随着模式识别技术的发展,光谱识别技术已成为医药、环保、石化等行业进行快速检测的重要的手段。神经网络具有非线性映射、自适应学习、鲁棒性和容错性等特点,在信号处理、知识工程、模式识别等领域有着广泛的应用。文章以符合朗伯-比尔定律的光谱信号为研究对象,概述了运用神经网络进行模式识别的基本原理,随后根据光谱识别的具体要求,提出了基于多特征和神经网络的光谱识别方案,并进行了系统设计,建立了基本的模型框架。最后运用实例对该方法进行了说明。     

Word recognition in noise at higher-than-normal levels: decreases in scores and increases in masking

Under certain conditions, speech recognition in noise decreases above conversational levels when signal-to-noise ratio is held constant. The current study was undertaken to determine if nonlinear growth of masking and the subsequent reduction in "effective" signal-to-noise ratio accounts for this decline. Nine young adults with normal hearing listened to monosyllabic words at three levels in each of three levels of a masker shaped to match the speech spectrum. An additional low-level noise equated audibility by producing equivalent masked thresholds for all subjects. If word recognition was determined entirely by signal-to-noise ratio and was independent of overall speech and masker levels, scores at a given signal-to-noise ratio should remain constant with increasing level. Masked pure-tone thresholds measured in the speech-shaped maskers increased linearly with increasing masker level at lower frequencies but nonlinearly at higher frequencies, consistent with nonlinear growth of upward spread of masking that followed the peaks in the spectrum of the speech-shaped masker. Word recognition declined significantly with increasing level when signal-to-noise ratio was held constant which was attributed to nonlinear growth of masking and reduced "effective" signal-to-noise ratio at high speech-shaped masker levels, as indicated by audibility estimates based on the Articulation Index.     

Speech recognition and the Articulation Index for normal and hearing-impaired listeners

The purpose of this experiment was to determine the applicability of the Articulation Index (AI) model for characterizing the speech recognition performance of listeners with mild-to-moderate hearing loss. Performance-intensity functions were obtained from five normal-hearing listeners and 11 hearing-impaired listeners using a closed-set nonsense syllable test for two frequency responses (uniform and high-frequency emphasis). For each listener, the fitting constant Q of the nonlinear transfer function relating AI and speech recognition was estimated. Results indicated that the function mapping AI onto performance was approximately the same for normal and hearing-impaired listeners with mild-to-moderate hearing loss and high speech recognition scores. For a hearing-impaired listener with poor speech recognition ability, the AI procedure was a poor predictor of performance. The AI procedure as presently used is inadequate for predicting performance of individuals with reduced speech recognition ability and should be used conservatively in applications predicting optimal or acceptable frequency response characteristics for hearing-aid amplification systems.     

基于最大似然多项式回归的鲁棒语音识别

本文针对最大似然线性回归算法线性假设的缺点,将多项式回归方法用于模型自适应,构建了基于最大似然多项式回归的非线性模型自适应算法。该算法在对数谱域用多项式回归方法,逼近每个Mel子带上识别环境模型均值与训练环境模型均值之间的非线性关系。多项式系数通过EM算法和最大似然准则从识别环境下的少量自适应数据中估计。实验结果表明,二阶多项式就可以较好地逼近模型均值的非线性环境变换关系。在噪声补偿和说话人自适应实验中,最大似然多项式回归算法的误识率都明显低于最大似然线性回归算法。本文算法较好地克服了线性模型自适应算法线性假设的缺陷,可同时减小噪声,和说话人的改变或其它因素对语音识别系统的影响,尤其适合说话人和噪声的联合自适应。      

基于微观本义复原的人脸图像光照正规化

为解决变光照下人脸识别的识别率低,光照正规化算法复杂.不易实现的问题,提出一个新的光照正规化方法一微观本义复原,即通过微观邻域上像素本义特征的定义,将整体图像上变光照下的非线性灰度变化转化为微观邻域内的线性变化,一定程度上避免了包括图像自身结构在内的不确定因素对图像复原的不利影响;并以邻域内的本义特征为光照不敏感特征,对本义特征进行结构编码,并用最小二乘法拟合编码值与光照方向之间的关系,最后根据得到的光照正规化参数复原图像.实验结果表明,该方法算法简单,易于实现,能适应实时的人脸识别系统,在光照变化90.以内的Yale B库的平均识别率可达94.1%.     

Maximum likelihood polynomial regression for robust speech recognition

The linear hypothesis is the main disadvantage of maximum likelihood linear regression (MLLR).This paper applies the polynomial regression method to model adaptation and establishes a nonlinear model adaptation algorithm using maximum likelihood polynomial regression(MLPR)for robust speech recognition.In this algorithm,the nonlinear relationship between training and testing Gaussian means in every Mel channel is approximated by a set of polynomials and the polynomial coefficients are estimated from adaptation data in test environment using the expectation-maximization(EM)algorithm and maximum likelihood(ML) criterion.The experimental results show that the second-order polynomial can approximate the actual nonlinear function better and in noise compensation and speaker adaptation,the word error rates of MLPR are significantly lower than those of MLLR.The proposed MLPR algorithm overcomes the limitation of linear hypothesis well and can decrease the impact of noise,speaker and other factors simultaneously.It is especially suitable for joint adaptation of speaker and noise.     

Nonlinear modeling technique for the analysis of DNA chains

Using a simple computational procedure, we examine DNA chains from different species in order to prove their nonlinear deterministic structures. This procedure applies a nonlinear modeling technique based upon quantitative comparison of the neighborhoods from similar DNA subsegments of size d. Our results reveal that noncoding regions exhibit a deterministic signature at sizes larger than a characteristic dimension d(c). Applications to evolutionary categories and recognition of different DNA regions are discussed.     

Automatic speech recognition using psychoacoustic models.

An approach to automatic speech recognition is described, which, in a straightforward way, follows the concept of (1) preprocessing in terms of auditory parameters and (2) subsequent classification and recognition. The preprocessing system has been realized in analog hardware, while recognition is carried out on a digital computer. In the preprocessing system, the essential psychoacoustic principles of the perception of loudness, pitch, roughness, and subjective duration are implemented with some approximation. The system essentially consists of 24 bandpass filters, nonlinear transformation of each filter output into specific loudness and specific roughness, and final transformation of these parameters into total loudness, total roughness, and three spectral momenta. As a means to further reduce the information flow, continuous selection of dominant parameters is also considered on the basis of psychoacoustic data. The subsequent recognition process is mainly characterized by (1) discrimination between speech and silent periods, (2) detection of syllable peaks and classification of syllable nuclei, and (3) assumption of syllable boundaries and classification of consonant clusters. Though the entire system as yet is far from being complete and perfect, the present results indicate that the concept provides a systematic and promising way towards automatic recognition of continuous speech.     

Spatially regularized and locality-constrained linear coding for human action recognition

To reduce quantization error, preserve the manifold of local features, distinguish the ambiguous features, and model the spatial configuration of features for Bag-of-Features (BoF) model-based human action recognition, a novel feature coding method called spatially regularized and locality-constrained linear coding (SLLC) is proposed. The spatial regularization and locality constraint are involved in the feature coding phase to model the spatial configuration of features and preserve their nonlinear manifold. The action recognition experimental results on benchmark datasets show that SLLC achieves better performance than the state-of-the-art feature coding methods such as soft vector quantization, sparse coding, and locality-constrained linear coding.     

Phase-only filtering based on sliced orthogonal nonlinear generalized decomposition

The sliced orthogonal nonlinear generalized (SONG) decomposition and correlation have been demonstrated to be powerful tools in digital image processing and promising for nonlinear optical information processing. In this paper, we propose an optical phase-only filtering system based on SONG decomposition (PBS), in which the phase-only filtering of the target and the input scene binary slices are performed separately by pairs and then added together. We numerically show that the PBS has extremely high and sharp output correlation peak compared with other optical correlators. Furthermore, such a SONG decomposition based phase-only filtering naturally inherits the nature of SONG decomposition, which has strong robustness to additive Gaussian noise and substitutive noise, and also the high light efficiency of phase-only filtering. We demonstrate that the PBS may serve as an optimized optical correlation scheme, which is promising in nonlinear optical pattern recognition.     

Detecting stop consonants in continuous speech

The problem of implementing a detector for stop consonants in continuously spoken speech is considered. The problem is posed as one of finding an optimal filter (linear or nonlinear) that operates on a particular appropriately chosen representation, and ideally outputs a 1 when a stop occurs and 0 otherwise. The performance of several variants of a canonical stop detector is discussed and its implications for human and machine speech recognition is considered.     

Phoneme discrimination using connectionist networks

The application of connectionist networks to speech recognition is assessed using a set of eight representative phonetic discrimination problems chose with respect to a theory of phonetics. A connectionist network model called the temporal flow model (TFM) is defined which represents temporal relationships using delay links and permits general patterns of connectivity. It is argued that the model has properties appropriate for time varying signals such as speech. Networks are trained using gradient descent methods of iterative nonlinear optimization to reduce the mean-squared error between the actual and the desired response of the output units. Separate network solutions are demonstrated for all eight phonetic discrimination problems for one male speaker. The network solutions are analyzed carefully and are shown in every case to make use of known acoustic phonetic cues. The network solutions vary in the degree to which they make use of context-dependent cues to achieve phoneme recognition. The network solutions were tested on data not used for training and achieved an average accuracy of 99.5%. It is concluded that acoustic phonetic speech recognition can be accomplished using connectionist networks.     

基于支持向量机的非线性荧光光谱的识别

提出将支持向量机网络应用于含不同浓度杂质气体的非线性荧光光谱的识别。由于原始光谱数据的光谱通道数目很大,首先用小波变换去噪压缩,然后采用主成分分析方法对光谱信息进行连续两次的特征提取。在保持原光谱数据主要信息基本不变的情况下,将数据维数由3979压缩到514(小波变换)并提取9个主成分。这样,不仅减少了网络的输入维数,而且加快了网络的训练速度。实验结果表明,无论对训练样本还是未学习过的测试样本,其正确识别率均可达到100%。网络的训练和测试速度较快,可以更有效地应用于大气杂质气体的实时监测。     

Weighted nonlinear correlation for controlled discrimination capability

We recently demonstrated the high discrimination capability as well as the high sensitivity to small intensity variations of the sliced orthogonal nonlinear generalized (SONG) correlation. This nonlinear correlation has a correlation matrix representation. Previous papers considered only the principal diagonal elements of the correlation matrix. We propose using the off-diagonal non-zero elements of the SONG correlation matrix in order to achieve variable discrimination performance and controlled detection adapted to the gray-scale variations. Moreover, we introduce negative coefficients in order to improve the discrimination properties of the SONG correlation. To control the degree of recognition we define a correlation that is related with the degree of degradation. The proposed changes to the SONG correlation may be implemented by means of optics.     

Surface-enhanced raman scattering and nonlinear optics applied to electrochemistry

The most recently developed diagnostic technique in metal-electrolyte and metal-gas interfaces adapts spontaneous Raman scattering and nonlinear optical generation, techniques normally applied to bulk media, to surface science investigation. For certain metallic surfaces, an enormous increase exists in the Raman (as much as 10⁶ to 10⁸ times) and nonlinear optical signals resulting from submonolayer coverage of molecular adsorbates at the interface. Spontaneous Raman scattering and nonlinear optical scattering are well developed in both theory and practice for the analysis of molecular structure and concentration in bulk media. Instrumentation to generate and detect these inelastically scattered signals is readily available and is adequate for adaption to surface science. However, the mechanism (or mechanisms) giving rise to such a large enhancement at the interfaces is still being actively researched and remains controversial. Theoretical and experimental investigations related to the underlying physics of this enhancement and the application of such surface enhancement as a vibrational probe for adsorbates on the metal surface have been labeled “surface-enhanced Raman scattering” (SERS) and “surface-enhanced nonlinear optics”. Soon after the recognition that molecules adsorbed onto metal electrodes under certain conditions exhibit an anomalously large Raman scattering efficiency,^1–3 it became evident that such a phenomenon makes possible an in situ diagnostic probe for detailed and unique vibrational signatures of adsorbates in the ambient phase (electrolyte and atmospheric gas surroundings). Optical spectroscopy in the visible range has a much higher energy resolution (e.g., 0. I cm-I) than is presently available in electron energy loss spectroscopy (EELS), as well as the capability to measure much lower frequency modes (e.g., as low as 5 cm^?1) than is possible in infrared spectroscopy. Perhaps the most significant attribute of SERS and surface-enhanced nonlinear optical scattering is that the surrounding media in front of the interface (e.g., several meters of gas and several centimeters of liquid) do not introduce optical loss or overwhelmingly large signals. The recognition that SERS is capable of performing vibrational spectroscopy with this resolution, frequency range, and in such dense surroundings has therefore brought an explosion of activity to the field since 1977.