基于局部保持投影与隐马尔可夫模型的维文字符识别

针对传统隐马尔可夫模型(HMM)在对手写维吾尔文字符建模时,字符宽度变化大,模型训练收敛缓慢,且易陷入局部极值的问题,提出一种基于保局投影(LPP)与HMM相结合的维吾尔字符识别方法。首先,通过高度归一化保持原图像的宽高比,用滑动窗获取子图像序列,形成观测向量序列;其次,采用局部保持投影将观测序列映射到低维空间,并用随机抽样方法降低邻接图矩阵的规模;最后,采用新观测序列训练HMM。该算法在降维的同时提高了HMM的收敛速度,降低了陷入局部极值的风险。实验结果显示,算法的平均收敛步数减少,错误率降低,表明算法是有效的。     

On-line recognition of cursive Korean characters using DP matching and fuzzy concept

This paper proposes an on-line recognition method of cursive Korean characters based on dynamic programming (DP) matching and fuzzy concept. The proposed algorithm, invariant to rotation and size, reduces greatly the computational requirement of DP matching by matching phonemes rather than character patterns, where the angle difference and the ratio of lengths between input and reference patterns are adopted as matching features. Correct matching of poorly-written cursive characters becomes possible by introducing the fuzzy concept in representing phoneme features and the positional relationships between adjacent phonemes. Computer simulation results show the effectiveness of the proposed algorithm.     

改进的HMM和小波神经网络的抗噪语音识别

通过MFFC计算出的语音特征系数,由于语音信号的动态性,帧之间有重叠,噪声的影响,使特征系数不能完全反映出语音的信息。提出一种隐马尔可夫模型（HMM）和小波神经网络（WNN）混合模型的抗噪语音识别方法。该方法对MFCC特征系数利用小波神经网络进行训练,得到新的MFCC特征系数。实验结果表明,在噪声环境下,该混合模型比单纯HMM具有更强的噪声鲁棒性,明显改善了语音识别系统的性能。     

脱机手写体签名识别的小波包隐马尔可夫模型

提出了一种基于小波包隐马尔可夫的脱机手写体签名识别方法。该方法用小波包对归一化的签名图像进行特征提取,用混合高斯模型刻画各频带的小波包的系数分布,并用隐马尔可夫的状态转移模型描述了高斯模型在各频带间的相关性和依赖性。该方法数据预处理简单,特征提取完全可逆,避免了复杂分割,很好地描述了签名图像的小波包分解的统计特性,实验表明具有较好的抗噪性、鲁棒性、适应性和较高的识别率。     

On-line recognition of hand-printed Korean characters

The Korean alphabet is a set of phonetic symbols which are combined to form characters, somewhat in the Chinese style. The phonetic quality of the symbols naturally limits the combinations that are useful, and character formation constraints further limit these combinations. Advantage of this is made in an on-line computer recognition system. Korean characters are entered by means of a graphic tablet, using the standard stroke sequences taught in schools. Recognition is perfect for carefully drawn characters and nearly so for characters written at an unhurried rate, provided the system is tuned to the writer's style.     

Zero knowledge hidden Markov model inference

Hidden Markov models (HMMs) are widely used in pattern recognition. HMM construction requires an initial model structure that is used as a starting point to estimate the model’s parameters. To construct a HMM without a priori knowledge of the structure, we use an approach developed by Crutchfield and Shalizi that requires only a sequence of observations and a maximum data window size. Values of the maximum data window size that are too small result in incorrect models being constructed. Values that are too large reduce the number of data samples that can be considered and exponentially increase the algorithm’s computational complexity. In this paper, we present a method for automatically inferring this parameter directly from training data as part of the model construction process. We present theoretical and experimental results that confirm the utility of the proposed extension.     

基于隐马尔可夫模型的运动目标轨迹识别 *

引入改进的隐马尔可夫模型算法,针对真实场景中运动目标轨迹的复杂程度对各个轨迹模式类建立相应的隐马尔可夫模型,利用训练样本训练模型得到可靠的模型参数;计算测试样本对于各个模型的最大似然概率,选取最大概率值对应的轨迹模式类作为轨迹识别的结果,对两种场景中聚类后的轨迹进行训练与识别。实验结果表明,平均识别率分别达到87.76 %和94. 19%。     

基于连续隐马尔可夫模型的人脸识别方法

提出了一种基于连续隐马尔可夫模型的人脸图像识别方法,主要内容包括以下方面:①由于奇异值向量具有稳定性.转置不变性等特点,对归一化的人脸图像,采用奇异值分解抽取人脸图像特征作为观察值序列;②在人脸识别中应用连续隐马尔可夫模型,采用双高斯概率密度函数训练,建立HMM模型,再利用建好的HMM模型进行识别.实验结果显示,所提出的方法减少了数据计算量,运行速度快,并提高了识别率,完全满足人脸识别系统实时性要求.     

Driving intention recognition and behaviour prediction based on a double-layer hidden Markov model

We propose a model structure with a double-layer hidden Markov model (HMM) to recognise driving intention and predict driving behaviour. The upper-layer multi-dimensional discrete HMM (MDHMM) in the double-layer HMM represents driving intention in a combined working case, constructed according to the driving behaviours in certain single working cases in the lower-layer multi-dimensional Gaussian HMM (MGHMM). The driving behaviours are recognised by manoeuvring the signals of the driver and vehicle state information, and the recognised results are sent to the upper-layer HMM to recognise driving intentions. Also, driving behaviours in the near future are predicted using the likelihood-maximum method. A real-time driving simulator test on the combined working cases showed that the double-layer HMM can recognise driving intention and predict driving behaviour accurately and efficiently. As a result, the model provides the basis for pre-warning and intervention of danger and improving comfort performance.     

基于加权HMM的车辆电源系统状态预测

针对车辆电源系统状态趋势问题,提出了一种加权隐马尔可夫模型的状态预测方法。通过建立电源系统的隐马尔可夫模型,利用加权预测思想对隐马尔可夫模型中隐状态序列进行预测,将最大概率隐状态利用观测概率密度计算出状态观测值。通过对电压调节脉宽信号的导通率进行预测,并与BP神经网络和自回归（AR）模型对相同序列的预测结果进行对比,结果表明该方法对系统的状态变化具有较好的预测能力。     

Handwritten word-spotting using hidden Markov models and universal vocabularies

Handwritten word-spotting is traditionally viewed as an image matching task between one or multiple query word-images and a set of candidate word-images in a database. This is a typical instance of the query-by-example paradigm. In this article, we introduce a statistical framework for the word-spotting problem which employs hidden Markov models (HMMs) to model keywords and a Gaussian mixture model (GMM) for score normalization. We explore the use of two types of HMMs for the word modeling part: continuous HMMs (C-HMMs) and semi-continuous HMMs (SC-HMMs), i.e. HMMs with a shared set of Gaussians. We show on a challenging multi-writer corpus that the proposed statistical framework is always superior to a traditional matching system which uses dynamic time warping (DTW) for word-image distance computation. A very important finding is that the SC-HMM is superior when labeled training data is scarce—as low as one sample per keyword—thanks to the prior information which can be incorporated in the shared set of Gaussians.     

基于隐马尔可夫模型的内存碎片检测机制

当前VxWorks操作系统缺少内存碎片的检测机制。通过增加内存统计信息,基于隐马尔可夫模型的检测程序分析出系统中哪些任务可能是造成内存碎片的根源。软件开发人员根据分析结果对可能造成内存碎片的代码进行优化,且优化前后的分析数据表明内存碎片问题得到了有效的改善,可满足嵌入式设备减少内存碎片的需求。     

基于轮廓特征的HMM手写数字识别

在手写数字识别中,边界链码和环构成了对字符轮廓的完整描述。针对手写数字的特点,建了24种笔划。首先将样本边界链码转化成由24个笔划组成的特征值,再加上环特征,构成整个特征值。然后利用隐马尔可夫模型（HMM）对提取的特征值进行分类识别。首次将字符轮廓特征应用在基于HMM的手写数字识别中,在识别MNIST字库上,取得了92.2%的识别率。     

A face recognition system based on Pseudo 2D HMM applied to neural network coefficients

Face recognition from an image or video sequences is emerging as an active research area with numerous commercial and law enforcement applications. In this paper different Pseudo 2-dimension Hidden Markov Models (HMMs) are introduced for a face recognition showing performances reasonably fast for binary images. The proposed P2-D HMMs are made up of five levels of states, one for each significant facial region in which the input frontal images are sequenced: forehead, eyes, nose, mouth and chin. Each of P2-D HMMs has been trained by coefficients of an artificial neural network used to compress a bitmap image in order to represent it with a number of coefficients that is smaller than the total number of pixels. All the P2-D HMMs, applied to the input set consisting of the Olivetti Research Laboratory face database combined to others photos, have achieved good rates of recognition and, in particular, the structure 3-6-6-6-3 has achieved a rate of recognition equal to 100%.     

基于隐马尔可夫模型的语音激活检测算法

针对现有基于隐马尔可夫模型（HMM）的语音激活检测（VAD）算法对噪声的跟踪性能不佳的问题,提出采用Baum-Welch算法对具有不同特性的噪声进行训练,并生成相应噪声模型,建立噪声库的方法。在语音激活检测时,根据待测语音背景噪声的不同,动态地匹配噪声库中的噪声模型;同时,为了适应语音信号的实时处理,降低了语音参数提取的复杂度,并对判决阈值提出改进,以保证语音信号帧间的相关性。在不同噪声环境下对改进算法进行性能测试并与自适应多速率编码（AMR）标准、国际电信联盟电信标准分局（ITU-T）的G.729B标准比较,测试结果表明,改进算法在实时语音信号处理中能够有效提高检测的准确率及噪声跟踪能力。