一种基于段级特征和自动标识的语言辨识算法

本文研究了一种结合"声学信息"和"音素配位学信息"进行语言辨识的新算法,首先在预处理中对语音进行自动分段,在特征层上引入带有长时信息的段级特征参数--段级移位差分倒谱,在模型层上利用高斯混合模型(Gaussi-an Mixture Model,GMM)将语音信号自动标识为符号序列,进而引入多元语言模型(Multi-gram Language Model,MLM)来对"音素配位学信息"进行建模,最后将"GMM得分"和"MLM得分"送入后端多分类支持向量机模型得到最终识别结果.相关实验表明,新系统不需手工标识的语料,识别速度快,对OGI标准语料库中的五种语言获得了开集正识率为78.84%的结果.     

应用说话人聚类技术改善语言辨识系统识别率

说话人差异是影响语言辨识系统性能的一个重要方面.采用说话人聚类技术对训练数据进行了预分类,以此为基础训练各种语言模型,得到聚类后的语言辨识系统.高斯混合模型、高斯混合模型-全局背景模型和遍历隐马尔可夫模型3种语言模型的实验证明,这种方法是有效的,它普遍地减小了说话人差异对语言辨识的影响,提高了语言辨识率.     

基于说话人聚类和高斯混合模型的语言辨识研究

本文给出了一种语言辨识的新方法。通常来讲,语言辨识系统是说话人无关的,但说话人的个体特征对语言辨识系统有很大的影响,文本采用了一种粗分类精识别的思想,利用说话人聚类技术有效解决了粗分类的问题,对每类相近说话人集合建立模型,然后进行识别。实验表明,该方法对于说话人无关的语言辨识问题是有效的。     

电力电子器件模型参数辨识及其有效性验证

准确的模型及模型参数是安全、合理使用电力电子器件的保证。本文采用了一种基于实验和仿真的电力电子器件模型参数辨识方法，通过优化算法，使仿真波形逐步向实验波形逼近，以达到参数辨识的目的．并在此基础上，对所辨识的参数进行了有效性验证。文中以PIN功率二极管BYT120-100为例．给出了模型参数辨识及有效性验证结果，并进一步论述了IGBT模型参数辨识及其有效性验证的基本思路。     

语言辨识系统的决策级融合研究

文中利用不同参数提取方法对语言辨识系统决策级融合进行了研究。文中给出4种融合方式，实验系统采用了GMM—UBM模型，利用OGI—TS多语种电话语音语料库，对决策级融合性能进行了评估。实验表明，利用决策级融合技术，选择最佳融合方式，可以很好地改善语言辨识率。     

汉语连续语音识别中不同基元声学模型的复合

该文研究由不同声学基元训练的声学模型的复合。在汉语连续语音识别中,流行的基元包括上下文相关的声韵母基元和音素基元。实验发现,有些汉语音节在声韵母模型下有更高的识别率,有些音节在音素模型下有更高的识别率。该文提出一种复合这两种声学模型的方法,一方面在识别过程中同时使用两种模型,另一方面在识别过程中避开造成低识别率的模型。实验表明,采用本文的方法后,音节错误率比音素模型和声韵母模型分别下降了9.60%和6.10%。     

基于GMM-UBM模型的语言辨识研究

与说话人识别、连续语音识别相比,自动语言辨识是一个相对较新的研究,而且是一项较难的课题。本文给出了一种基于GMM-UBM模型的语言辨识系统,并利用OGI-TS电话语音库对算法的性能进行了测试,然后给出了实验结果。实验结果表明,该算法也是进行语言辨识的一种有效方法。     

用数字信号处理方法研究汉语普通话音素的区别特征

本文运用语言信号数字处理方法，研究了汉语普通话音素的区别特征，研究结果进一步完善了汉语普通话音素的区别特征矩阵表，将为基于音素的计算机汉语普通话语音分析、合成和识别提供了一种有效的参考方法。     

IGBT模型参数辨识及其有效性验证

由于导通及开关损耗小以及易于使用等优点，IGBT在电力电子系统中得到越来越广泛的应用。在设计电路之前，需要精确的器件模型及模型参数对电路进行仿真。本文提出一种基于实验测量、仿真及优化算法的IGBT模型参数辨识方法。以BUP302为例，给出了静态参数及动态参数的结果。在参数辨识的基础上，文中还提出了模型参数有效性验证的方法，最后给出了有效性验证结果。     

非语言符号在访谈节目中的运用

在传播学中将一切传递信息的形式统称为符号,符号本身具有传递信息的功能,据传播学对符号的划分,分为语言符号和非语言符号.语言符号是人类交流的主要工具,但非语言符号也起着不可或缺的作用,它起到强调、补充、辅助、代替、调节信息等作用,尤其是在以交际为核心的访谈节目中,语言符号虽然是主要传播手段,但与非语言符号相结合才能够形成完整的表意体系.本文以访谈节目中非语言符号的使用为切入口,探索思考非语言符号当中的身体非语言符号和物体非语言符号在访谈节目中的具体使用技巧.     

XML Based UIScript

1　Introduction　Intelligentnetwork[1～2 ] wasintroducedinthe1 980 s.Itisbasedontheimplementationofaser vice independentandmachine independentarchitec tureonwhichnetworkserviceproviderscancreatenewservicesforcustomers.Thebasicelementsinintelligentnetworkare :theServiceSwitchingPoint(SSP)whichdetectsrequestsforINservice ;theServiceControlPoint (SCP)whichcontainsservicelogictoprovideINservices;theIntelligentPeriph eral (IP)whichprovidesspecialresources;andtheServiceManagementPoint (SM…     

The Z specification dependence graph

The lack of existing solutions makes it really hard to understand formal specification languages since the application domain for representations is useful for the purpose of carrying out certain software engineering operations such as slicing and the computation of program metrics.A Z specification dependence graph is presented in this letter. It draws on the strengths of a range of earlier works and adapts them, if necessary, to the Z language.     

Reviewing automatic language identification

The Oregon Graduate Institute Multi-language Telephone Speech Corpus (OGI-TS) was designed specifically for language identification research. It currently consists of spontaneous and fixed-vocabulary utterances in 11 languages: English, Farsi, French, German, Hindi, Japanese, Korean, Mandarin, Spanish, Tamil, and Vietnamese. These utterances were produced by 90 native speakers in each language over real telephone lines. Language identification is related to speaker-independent speech recognition and speaker identification in several interesting ways. It is therefore not surprising that many of the recent developments in language identification can be related to developments in those two fields. We review some of the more important recent approaches to language identification against the background of successes in speaker and speech recognition. In particular, we demonstrate how approaches to language identification based on acoustic modeling and language modeling, respectively, are similar to algorithms used in speaker-independent continuous speech recognition. Thereafter, prosodic and duration-based information sources are studied. We then review an approach to language identification that draws heavily on speaker identification. Finally, the performance of some representative algorithms is reported     

基于TMS320C25音频处理系统的C语言编程

这篇文章着重讨论了Ｃ语言程序在以ＴＭＳ３２０Ｃ２５为基础的处理系统中的开发和应用。详细介绍了从Ｃ源程序到适用于Ｃ２５格式的程序的开发过程，同时讨论了在开发过程中可能遇到的问题及其解决办法。同时，文章对两种编程语言在ＴＭＳ３２０Ｃ２５系统中的应用前景作了简单的比较和预测。     

Multilevel Location Trigger in Distributed Mobile Environments for Location‐Based Services

In location‐based services, most trigger technologies have been implemented on the server side by periodically requesting the locations of mobile phones from mobile network servers. However, bottlenecks and communication interruptions occur when the servers are overloaded by trigger requests. In this letter, we propose a new multilevel location trigger specification which distributes the event detecting role to mobile phones and redesigns the location triggering into a multilevel step. Our suggested location trigger specification can reduce bottlenecks caused by triggers in a mobile core network and reduce power consumption caused by embedded GPS devices in mobile phones.     

Inayan/nga-ag and other indigenous codes: How the Applai and Bontok Igorot’s indigeneity found its way into the mobile world

Scholars recognize that the mobile phone is so pervasive that studying it in all context is a necessary project. Very few literature on mobile phone use in indigenous communities in the Philippines has been written. Currently, there are no indigeneity-segregated data on mobile phone use in the country. Even mobile phone use in rural Philippines remains understudied. The country is home to 110 groups of indigenous peoples (IP) with 126 distinct languages and very diverse traditions and practices.The Cordillera mountain range in the northern part of the Philippines is recognized as the traditional domain of the Igorot, or people of the mountains. Its geopolitical location, the Cordillera Administrative Region, has one of the densest populations of indigenous peoples in the country. Under the label Igorot are a number of indigenous communities that has embraced the label as a form of resistance.The paper focuses on three indigenous communities in Mountain Province: (1) Barangay Laylaya, Besao, an Applai community that speaks the Western Kankana-ey language, the farthest with only one transport vehicle that leaves the village very early in the morning and returns late in the afternoon; (2) Payag-eo, Suyo, an Applai community that also speaks the Western Kankana-ey language; politically part of the tourist destination Sagada but is completely cut-off in terms of transport accessibility and is closer to the Bontok tribe geographically; and (3) Guina-ang, Bontoc, with people from the Bontok tribe who speaks iGuina-ang, one of the variations of the Bontok language, the village is an hour away from the provincial center where branches of national offices are housed and major trade activities take place.Guided by the culture-centered approach, an alternative lens built on the notion that understanding and negotiating meanings are embedded within cultural contexts and values, the paper looks into how mobile phones has affected, mediated, manifested and complicated the indigeneity of the Igorot. Specifically, it discusses how the indigenous social code inayan/nga-ag permeated the community’s use of their mobile phones. From simple “Do not bring your mobile phones on the table.” to more morally imposing “Do not use your mobile phone for indecent and illicit acts.” This concept of inayan remains integral to the Igorot culture and is now intertwined with their use of their mobile phones. And as the Applai and Bontok Igorot struggle to keep their lifeways and traditions intact, the mobile phone continues to both facilitate this and at the same time get in the way.     

上下文无关文法与无限状态自动机

目前，在研究上下文无关语言时常用的形式系统是上下文无关文法和下推自动机，在研究正则语言时常用的形式系统是正则文法和有限状态自动机。正则文法中的符号和有限状态自动机的符号之间的对应关系比较明显，因此，两种系统之间的转换比较容易，并且在这两种系统中观察语言时，可以得到一致的解释。