语音的

Reason is such, because we think to did not come in speech calculative, the browser is the platform of a special core, because later of the user get online, the osculatory speech computation of the user, no matter be in PC, mobile phone or over the hardware of other, it needs a very powerful browser.

理由是这样的，因为我们认为在语音计算的未来，浏览器是一个非常核心的平台，因为以后用户的上网、用户的接触语音计算，无论是在PC、手机或者是其它的硬件之上，它需要一个非常强而有力的浏览器。

In accordance with chaotic essence of speech signals , syllable segmentation in continuous speech is researched by fractal theory . an approach of syllable segmentation using variance fractal dimension is proposed , its performance is analyzed in detail . the method can discriminate between voiced and unvoiced , between surd and sonant , but it can hardly discriminate between sonant

本文根据语音信号的混沌本质，利用分形理论研究了汉语连续语音中的音节分割问题，提出了基于方差分形维数的音节分割方法，并详细分析了该方法的性能，它能很好地解决了无声与有声、浊音与清音间的分割问题，但很难解决浊音间的分割问题，当浊音相连且过渡段较短时，该方法无法实现它们之间的分割。

The independence of the function is discussed focusing on the connection between man and language by analyzing the important concept of Chinese classical poetics Feelings-expressed-by-sound , which shows the opening and harmonious connection between the inner formalized feelings of man and the outer formalized sound of poetry or between man and language. As the formalized emotion, the feelings of the sound formed don't depend on social emotion.

生成的独立性：诗歌语音结构的生成具有"物之自则"即和谐自组织的的特性，也即不是在外力作用下而是按自身内在构型规律成型的；从人与语言的关系来看，古人用"体""用"并通过对体用一如的强调，来揭示人的内在性情与诗之"声"的高度交融关系，正是在"情之自溢"、"触物成声"的生成过程中，诗歌才做到体用一如——诗歌语音和谐构型的独立性离不开人的内在情文构型的独立性。

In order to decrease the error rate of speech recognition, speaker adaptation techniques are often used to adjust speaker-dependent acoustic models. MLLR (Maximum Likelihood Linear Regression) and MAP are two of the most popular techniques in recent years. MLLR uses the technique of regression trees. It calculates the transform matrix for each leaf node of the tree. This makes it possible to use fewer sentences to decrease the error rate of HMM-based speech recognition. However, while we examined the recognition result, we found that although the overall error rate decreased, but the error rate of certain confusable phones was higher. In order to solve this problem, we propose the use MPE (Minimum Phone Error Discriminative Training) to solve this problem.

在语音辨识技术的实际应用上，语者调适技术常被用做调整语者无关之声学模型，使其对於特定语者之辨识率有所提升，常见的语者调适技术有最大化相似度线性回归（Maximum Likelihood Linear Regression），其精神在於透过对语音辨识模型相近的mixture做群聚，再对各个群聚做调整，以期能达到利用少量语料，提升辨识率的目的，但其缺点为，在发音上相近的模型，其mixture原本就极为近似，若再分类为同一群聚做调整，容易因为使用者的发音习惯，而使模型偏向ㄓ或ㄗ，因此虽然整体的辨识率获得提升，但却造成混淆音的错误率上升。

Features Include:-Intelligent Voice Dictation with rapid correction ability -Document Player that can convert any document into audible files for later playback -Default and Creatable Voice Shortcuts that run programs, open documents or lead to Websites at your Command -Audible Announcement of OS messages -Multiple Voices to choose from -Continuous Evolution: Constant free updates to registered users.

功能包括：智能语音听写与迅速纠正的能力，文件播放器，可以转换成任何文件声音文件供以后播放，违约和Creatable语音快捷的运行程序，打开文件或导致网站在您的命令声短片的OS邮件，多选择的声音从连续进化：免费的恒定更新的注册用户。

We present an improved DP-MBE SAM; secondly, un-voice analysis is introduced to the DP-MBESAM, so that the improved ASA system can be used not only to separate concurrent vowels but concurrent speech signals with unvoice as well.

针对基于DP-MBE SAM的混叠浊音听觉场景分析系统存在的问题，提出了改进的基于DP-MBE SAM的混叠语音听觉场景分析系统，包括两部分内容，第一，针对DP-MBE SAM在实际的应用中存在的求解矩阵奇异引起的参数模糊及混叠信号谐波频率相同引起的参数模糊问题，借鉴多帧内插方法，提出了改进的DP-MBE SAM；第二，在基于DP-MBE SAM的混叠浊音听觉场景分析系统加入清音分离环节，将系统的应用从浊音推广到含有清浊音的语音信号。

The first idea to deal with the packet loss is the new proposal of multitask ATM cells and the required new ADPCM encoder. This scheme combines three tasks together, namely, reducing cell packeting delay, avoiding breakdown of speech decoders due to one lost cell, compensating or replacing the lost information segment, by dividing the cell payload field into two parts——main and additional information fields, the latter again consists of synchronization information and the coded speech information contained in the last cell.

首先以ATM网络为例，提出了新的信息分组多任务装配法，以及对应此方法的新型ADPCM语音编码系统，这种方法中，利用特殊设计的编码器，使ATM信元负载信息域，由主要和附加用户信息两部分组成，后者又由同步信息和前一个信元语音段信息组成，通过选择和调节信元中两部分信息域的比例，实现了小的分组装配时延、高的分组丢失补偿质量、强的阻断误差传递功能三者的结合。

Through the SPCE061A programming, this design requests us use the single chip microcomputer, PLC or other components to control the LED digital tube and the double color LED lamp on the traffic light control panel, enable the system to have the following functions: The four intersections of the crossroad all have double color LED (gives out red light and green light, superimposes for yellow light) to demonstrate the condition of every channel on the intersections; Each intersection has two digital tubes to demonstrate current traffic light countdown; use the button or other way to simulate the counting of the traffic flow volume.

设计要求通过SPCE061A的编程，使用单片机、PLC或其他器件来控制交通灯控制板上的LED数码管以及双色LED灯，使系统具有以下功能：十字路口的四个路口都有双色 LED （发红、绿光，叠加为黄色光）显示各路口的各条通道状态；每个路口有两位数码管显示当前红绿灯的倒计时；可利用按键或其它的方式模拟对车流量的计数；有四种工作状态：正常工作状态、紧急工作状态、设置状态和查询状态；红绿灯变换时有语音提示；查询时，可语音播报当前路况。

The second part describes the phonology of Sumitang Qidouhua including its consonants, vowels and tone, analyzes the rules of Wen and Bai, the difference between the old and the middle age, and offers the syllabary.The third part explores the close relationship between Sumitang Qidouhua and Songbai Tuhua by comparing the two kinds of Tuhua with Chengguan Tuhua in Jiangyong county, and puts forward a conclusion that the three kinds of Tuhua maybe come from the same origin although some difference exist among them. The comparison of the three kinds of Tuhua with the neighboring dialects testifies the above opinion. The forth part is the surplus of the paper.

第二部分：江华粟米塘七都话的语音系统这一部分全面分析了粟米塘七都话的声韵调及其语音特点、文白异读规律和老壮差异的例子，记录并整理了同音字表第三部分：江华粟米塘七都话与周边土话的比较这一部分首先将粟米塘七都话与江永松柏土话、城关土话进行了声韵调的全面比较，揭示了粟米塘七都话和松柏土话的密切联系，并推论这三种土话尽管存在各类差异，但应该有着相同的来源；然后再将这三种土话与周边土话进行比较，论证了上述观点。

At first, in this paper the background and basic principle of VoIP are introduced; the much popular VoIP technology based on H.323 protocol is studied; based on the function and main working process of IP telephone system, the principle of coding of compression and transfer technology of the real-time voice are analyzed; depended on the existed experimental environment, the program of voice part of PC to PC video telephone terminal based on LAN is designed and implemented; and the problems encountered during the design of program have been analyzed and solved.

本文首先介绍了IP电话的发展简史，IP电话技术原理；对目前最为流行的基于H.323协议的VoIP技术进行了研究；根据IP电话系统的功能和主要实现过程，分析实时语音压缩编码技术和传输技术原理；利用实验室现有的实验环境，设计并基本完成了一种基于局域网的PC to PC可视电话语音部分的编程；并对程序设计过程中所遇到的问题进行了分析和解决。

However, as the name（read－only memory）implies, CD disks cannot be written onorchanged in any way.

然而，正如其名字所指出的那样，CD盘不能写，也不能用任何方式改变其内容。

Galvanizes steel pallet is mainly export which suits standard packing of European Union, the North America. galvanizes steel pallet is suitable to heavy rack. Pallet surface can design plate type, corrugated and the gap form, satisfies the different requirements.

镀锌钢托盘多用于出口，替代木托盘，免薰蒸，符合欧盟、北美各国对出口货物包装材料的法令要求；喷涂钢托盘适用于重载上货架之用，托盘表面根据需要制作成平板状、波纹状及间隔形式，满足不同的使用要求。