A Modified Fixed‐Threshold SMO for 1‐Slack Structural SVMs

In this paper, we describe a modified fixed‐threshold sequential minimal optimization (FSMO) for 1‐slack structural support vector machine (SVM) problems. Because the modified FSMO uses the fact that the formulation of 1‐slack structural SVMs has no bias, it breaks down the quadratic programming (QP) problems of 1‐slack structural SVMs into a series of smallest QP problems, each involving only one variable. For various test sets, the modified FSMO is as accurate as existing structural SVM implementations (n‐slack and 1‐slack SVM‐struct) but is faster on large data sets.     

Kernel-based adversarial attacks and defenses on support vector classification

While malicious samples are widely found in many application fields of machine learning, suitable countermeasures have been investigated in the field of adversarial machine learning. Due to the importance and popularity of Support Vector Machines (SVMs), we first describe the evasion attack against SVM classification and then propose a defense strategy in this paper. The evasion attack utilizes the classification surface of SVM to iteratively find the minimal perturbations that mislead the nonlinear classifier. Specially, we propose what is called a vulnerability function to measure the vulnerability of the SVM classifiers. Utilizing this vulnerability function, we put forward an effective defense strategy based on the kernel optimization of SVMs with Gaussian kernel against the evasion attack. Our defense method is verified to be very effective on the benchmark datasets, and the SVM classifier becomes more robust after using our kernel optimization scheme.     

基于支持向量机的说话人辨认研究

支持向量机是统计学理论的一个重要的学习方法，也是解决模式识别问题的一个有力的工具，本文提出了用支持向量机来解决说话人辨认问题。结合语音信号的特点，解决了大数据量情况下支持向量机的训练问题。支持向量机对两类的分类问题有着突出的优势，本文用两种判决规则将两类问题应用到多类的识别问题。用支持向量机实现了一个与文本无关的说话人辨认系统，实验表明，本方法有良好的效果。     

基于支持向量机的异常值检测的两种方法

支持向量机逐渐成为机器学习的一种方法。异常值检测是支持向量机中一种特殊的分类问题,被称为一类分类。一类分类通过核映射确定一个包含正类样本的紧致区域,以便使异常值更容易暴露出来。介绍了一些一类分类算法的基本思想。     

Smooth support vector machine based on piecewise function

Support vector machines （SVMs） have shown remarkable success in many applications. However, the non-smooth feature of objective function is a limitation in practical application of SVMs. To overcome this disadvantage, a twice continuously differentiable piecewise-smooth function is constructed to smooth the objective function of unconstrained support vector machine （SVM）, and it issues a piecewise-smooth support vector machine （PWESSVM）. Comparing to the other smooth approximation functions, the smooth precision has an obvious improvement. The theoretical analysis shows PWESSVM is globally convergent. Numerical results and comparisons demonstrate the classification performance of our algorithm is better than other competitive baselines.     

The Philips Research system for continuous-speech recognition

This paper gives an overview of the Philips Research system for continuous-speech recognition. The recognition architecture is based on an integrated statistical approach. The system has been successfully applied to various tasks in American English and German, ranging from small vocabulary tasks to very large vocabulary tasks and from recognition only to speech understanding. Here, we concentrate on phoneme-based continuous-speech recognition for large vocabulary recognition as used for dictation, which covers a significant part of our research work on speech recognition. We describe this task and report on experimental results. In order to allow a comparison with the performance of other systems, a section with an evaluation on the standard North American Business news (NAB²) task (dictation of American English newspaper text) is supplied.     

Application of Emotion Recognition and Modification for Emotional Telugu Speech Recognition

Majority of the automatic speech recognition systems (ASR) are trained with neutral speech and the performance of these systems are affected due to the presence of emotional content in the speech. The recognition of these emotions in human speech is considered to be the crucial aspect of human-machine interaction. The combined spectral and differenced prosody features are considered for the task of the emotion recognition in the first stage. The task of emotion recognition does not serve the sole purpose of improvement in the performance of an ASR system. Based on the recognized emotions from the input speech, the corresponding adapted emotive ASR model is selected for the evaluation in the second stage. This adapted emotive ASR model is built using the existing neutral and synthetically generated emotive speech using prosody modification method. In this work, the importance of emotion recognition block at the front-end along with the emotive speech adaptation to the ASR system models were studied. The speech samples from IIIT-H Telugu speech corpus were considered for building the large vocabulary ASR systems. The emotional speech samples from IITKGP-SESC Telugu corpus were used for the evaluation. The adapted emotive speech models have yielded better performance over the existing neutral speech models.

     

Speaker Adaptation of Hybrid NN/HMM Model for Speech Recognition Based on Singular Value Decomposition

Recently several speaker adaptation methods have been proposed for deep neural network (DNN) in many large vocabulary continuous speech recognition (LVCSR) tasks. However, only a few methods rely on tuning the connection weights in trained DNNs directly to optimize system performance since it is very prone to over-fitting especially when some class labels are missing in the adaptation data. In this paper, we propose a new speaker adaptation method for the hybrid NN/HMM speech recognition model based on singular value decomposition (SVD). We apply SVD on the weight matrices in trained DNNs and then tune rectangular diagonal matrices with the adaptation data. This alleviates the over-fitting problem via updating the weight matrices slightly by only modifying the singular values. We evaluate the proposed adaptation method in two standard speech recognition tasks, namely TIMIT phone recognition and large vocabulary speech recognition in the Switchboard task. Experimental results have shown that it is effective to adapt large DNN models using only a small amount of adaptation data. For example, recognition results in the Switchboard task have shown that the proposed SVD-based adaptation method may achieve up to 3-6 % relative error reduction using only a few dozens of adaptation utterances per speaker.     

基于负相关学习的支持向量机集成算法

提出一种基于负相关学习理论的支持向量机集成方法，该方法能有效解除各支持向量机之问的相关性，使得集成个体有较大的差异。并采用了演化策略对支持向量机的核函数和相关参数白适应地进行选择。仿真实验表明．该方法不仅能有效地解决支持向量机模型选择的难题，而且能以很小的代价显著提高学习系统的泛化能力。     

An accurate traffic classification model based on support vector machines

Network traffic classification is a fundamental research topic on high‐performance network protocol design and network operation management. Compared with other state‐of‐the‐art studies done on the network traffic classification, machine learning (ML) methods are more flexible and intelligent, which can automatically search for and describe useful structural patterns in a supplied traffic dataset. As a typical ML method, support vector machines (SVMs) based on statistical theory has high classification accuracy and stability. However, the performance of SVM classifier can be severely affected by the data scale, feature dimension, and parameters of the classifier. In this paper, a real‐time accurate SVM training model named SPP‐SVM is proposed. An SPP‐SVM is deducted from the scaling dataset and employs principal component analysis (PCA) to extract data features and verify its relevant traffic features obtained from PCA. By employing PCA algorithm to do the dimension extraction, SPP‐SVM confirms the critical component features, reduces the redundancy among them, and lowers the original feature dimension so as to reduce the over fitting and increase its generalization effectively. The optimal working parameters of kernel function used in SPP‐SVM are derived automatically from improved particle swarm optimization algorithm, which will optimize the global solution and make its inertia weight coefficient adaptive without searching for the parameters in a wide range, traversing all the parameter points in the grid and adjusting steps gradually. The performance of its two‐ and multi‐class classifiers is proved over 2 sets of traffic traces, coming from different topological points on the Internet. Experiments show that the SPP‐SVM's two‐ and multi‐class classifiers are superior to the typical supervised ML algorithms and performs significantly better than traditional SVM in classification accuracy, dimension, and elapsed time.     

基于CNN-SVM性别组合分类的单通道语音分离

实际语音分离时,混合语音的说话人性别组合相关信息往往是未知的。若直接在普适的模型上进行分离,语音分离效果欠佳。为了更好地进行语音分离,本文提出一种基于卷积神经网络-支持向量机（CNN-SVM）的性别组合判别模型,来确定混合语音的两个说话人是男-男、男-女还是女-女组合,以便选用相应性别组合的分离模型进行语音分离。为了弥补传统单一特征表征性别组合信息不足的问题,本文提出一种挖掘深度融合特征的策略,使分类特征包含更多性别组合类别的信息。本文的基于CNN-SVM性别组合分类的单通道语音分离方法,首先使用卷积神经网络挖掘梅尔频率倒谱系数和滤波器组特征的深度特征,融合这两种深度特征作为性别组合的分类特征,然后利用支持向量机对混合语音性别组合进行识别,最后选择对应性别组合的深度神经网络/卷积神经网络（DNN/CNN）模型进行语音分离。实验结果表明,与传统的单一特征相比,本文所提的深度融合特征可以有效提高混合语音性别组合的识别率;本文所提的语音分离方法在主观语音质量评估（PESQ）、短时客观可懂度（STOI）、信号失真比（SDR）指标上均优于普适的语音分离模型。     

Support Vector Machines for Automated Recognition of Obstructive Sleep Apnea Syndrome From ECG Recordings

Obstructive sleep apnea syndrome (OSAS) is associated with cardiovascular morbidity as well as excessive daytime sleepiness and poor quality of life. In this study, we apply a machine learning technique [support vector machines (SVMs)] for automated recognition of OSAS types from their nocturnal ECG recordings. A total of 125 sets of nocturnal ECG recordings acquired from normal subjects (OSAS- ) and subjects with OSAS (OSAS+), each of approximately 8 h in duration, were analyzed. Features extracted from successive wavelet coefficient levels after wavelet decomposition of signals due to heart rate variability (HRV) from RR intervals and ECG-derived respiration (EDR) from R waves of QRS amplitudes were used as inputs to the SVMs to recognize OSAS +/- subjects. Using leave-one-out technique, the maximum accuracy of classification for 83 training sets was found to be 100% for SVMs using a subset of selected combination of HRV and EDR features. Independent test results on 42 subjects showed that it correctly recognized 24 out of 26 OSAS + subjects and 15 out of 16 OSAS - subjects (accuracy = 92.85%; Cohen's kappa value of 0.85). For estimating the relative severity of OSAS, the posterior probabilities of SVM outputs were calculated and compared with respective apnea/hypopnea index. These results suggest superior performance of SVMs in OSAS recognition supported by wavelet-based features of ECG. The results demonstrate considerable potential in applying SVMs in an ECG-based screening device that can aid a sleep specialist in the initial assessment of patients with suspected OSAS.     

Support Vector Machine Training for Improved Hidden Markov Modeling

We present a discriminative training algorithm, that uses support vector machines (SVMs), to improve the classification of discrete and continuous output probability hidden Markov models (HMMs). The algorithm uses a set of maximum-likelihood (ML) trained HMM models as a baseline system, and an SVM training scheme to rescore the results of the baseline HMMs. It turns out that the rescoring model can be represented as an unnormalized HMM. We describe two algorithms for training the unnormalized HMM models for both the discrete and continuous cases. One of the algorithms results in a single set of unnormalized HMMs that can be used in the standard recognition procedure (the Viterbi recognizer), as if they were plain HMMs. We use a toy problem and an isolated noisy digit recognition task to compare our new method to standard ML training. Our experiments show that SVM rescoring of hidden Markov models typically reduces the error rate significantly compared to standard ML training.     

语音技术在塔台模拟机上的应用

基于空管模拟训练机对空管人员训练的重要性和优越性以及塔席特点,利用计算机实现飞行器动态特征的场景再现,设定多场景、多任务的训练科目是目前采用的普遍做法。文章依托塔台模拟机性能特点,采用有限词汇量识别方法,结合空管发音规则、语法结构、发音标准以及中英双语等特点,采用动态时间归正识别技术和语音同步叠加算法实现语音编码、识别与合成,实现了多局部最优化决策。     

A prediction fusion method for reconstructing spatial temporal dynamics using support vector machines

In this paper, we propose a new spatial temporal predictor using support vector machine (SVM) and data fusion technique. SVMs are used as temporal predictors at different spatial domains and spatial temporal prediction is achieved by prediction fusion. Our proposed prediction fusion technique improves the prediction accuracy even in a non-Gaussian environment. The performance of the proposed spatial temporal predictor is analyzed. Based on real-life radar data, the proposed spatial temporal approach is shown to provide a more accurate model for sea-clutter data than the conventional methods.     

Localization Algorithm for Large Scale Wireless Sensor Networks Based on Fast-SVM

Sensor node localization is one of research hotspots in the applications of wireless sensor networks (WSNs) field. In recent years, many scholars proposed some localization algorithms based on machine learning, especially support vector machine (SVM). Localization algorithms based on SVM have good performance without pairwise distance measurements and special assisting devices. But if detection area is too wide and the scale of wireless sensor network is too large, the each sensor node needs to be classified many times to locate by SVMs, and the location time is too long. It is not suitable for the places of high real-time requirements. To solve this problem, a localization algorithm based on fast-SVM for large scale WSNs is proposed in this paper. The proposed fast-SVM constructs the minimum spanning by introducing the similarity measure and divided the support vectors into groups according to the maximum similarity in feature space. Each group support vectors is replaced by linear combination of “determinant factor” and “adjusting factor” which are decided by similarity. Because the support vectors are simplified by the fast-SVM, the speed of classification is evidently improved. Through the simulations, the performance of localization based on fast-SVM is evaluated. The results prove that the localization time is reduce about 48 % than existing localization algorithm based on SVM, and loss of the localization precision is very small. Moreover, fast-SVM localization algorithm also addresses the border problem and coverage hole problem effectively. Finally, the limitation of the proposed localization algorithm is discussed and future work is present.     

汉语大词汇量连续语音识别系统研究进展

本文综述了近年来大词汇量连续语音识别技术进步和发展,描述了大词汇量连续汉语语音识别系统的设计方法.对语音识别系统中的一些关键技术和原理进行了详细地分析和讨论,并对语音识别技术进一步发展中存在的问题和近年语音识别研究发展动向进行了讨论.     

基于人工蜂群算法的支持向量机参数优化及应用

为了解决常用的支持向量机(SVM)参数优化方法在寻优过程不同程度的陷入局部最优解的问题,提出一种基于人工蜂群(ABC)算法的SVM参数优化方法。将SVM的惩罚因子和核函数参数作为食物源位置,分类正确率作为适应度,利用ABC算法寻找适应度最高的食物源位置。利用4个标准数据集,将其与遗传(GA)算法、蚁群(ACO)算法、标准粒子群(PSO)算法优化的SVM进行性能比较,结果表明,本文方法能克服局部最优解,获得更高的分类正确率,并在小数目分类问题上有效降低运行时间。将本文方法运用到计算机笔迹鉴别,对提取的笔迹特征进行分类,与GA算法、ACO算法、PSO算法优化的SVM相比,得到了更高的分类正确率。     

Multitraining support vector machine for image retrieval.

Relevance feedback (RF) schemes based on support vector machines (SVMs) have been widely used in content-based image retrieval (CBIR). However, the performance of SVM-based RF approaches is often poor when the number of labeled feedback samples is small. This is mainly due to 1) the SVM classifier being unstable for small-size training sets because its optimal hyper plane is too sensitive to the training examples; and 2) the kernel method being ineffective because the feature dimension is much greater than the size of the training samples. In this paper, we develop a new machine learning technique, multitraining SVM (MTSVM), which combines the merits of the cotraining technique and a random sampling method in the feature space. Based on the proposed MTSVM algorithm, the above two problems can be mitigated. Experiments are carried out on a large image set of some 20,000 images, and the preliminary results demonstrate that the developed method consistently improves the performance over conventional SVM-based RFs in terms of precision and standard deviation, which are used to evaluate the effectiveness and robustness of a RF algorithm, respectively.     

An Eigen-Normal Approach for 3D Mesh Watermarking Using Support Vector Machines

The use of support vector machines （SVM） for watermarking of 3D mesh models is investigated. SVMs have been widely explored for images, audio, and video watermarking but to date the potential of SVMs has not been explored in the 3D watermarking domain. The proposed approach utilizes SVM as a binary classifier for the selection of vertices for watermark embedding. The SVM is trained with feature vectors derived from the angular difference between the eigen normal and surface normals of a 1-ring neighborhood of vertices taken from normalized 3D mesh models. The SVM learns to classify vertices as appropriate or inappropriate candidates for modification in order to accommodate the watermark. Experimental results verify that the proposed algorithm is imperceptible and robust against attacks such as mesh smoothing, cropping and noise addition.