公开(公告)号：US20020198712A1

公开(公告)日：2002-12-26

申请号：US10165973

申请日：2002-06-11

申请人： HEWLETT PACKARD COMPANY

发明人： Stephen John Hinde ,  Guillaume Belrose

IPC分类号： G10L015/00

CPC分类号： G10L13/02 ,  G10L15/06

摘要： A method is provided of generating an artificial language for use, for example, in human speech interfaces to devices. In a preferred implementation, the language generation method involves using a genetic algorithm to evolve a population of individuals over a plurality of generations, the individuals forming or being used to form candidate artificial-language words. The method is carried in a manner favouring the production of artificial-language words which are more easily correctly recognised by a speech recognition system and have a familiarity to a human user. This is achieved, for example, by selecting words for evolution on the basis of an evaluation carried out using a fitness function that takes account both of correct recognition of candidate words when spoken to a speech recognition system, and the similarity of candidate words to words in a set of user-favourite words.

摘要翻译： 提供了一种生成人造语言的方法，例如在与设备的人类语音界面中使用。 在优选实现中，语言生成方法涉及使用遗传算法来演化多代人群体，形成或被用于形成候选人工语言单词的个体。 该方法以有利于生产更容易被语音识别系统正确识别并且熟悉人类用户的人造语言词语的方式进行。 这是通过例如通过使用考虑到对语音识别系统进行口语的候选词的正确识别的适应度函数进行的评价来选择用于进化的单词，以及候选词与词语的相似性 在一组用户喜爱的单词中。

公开(公告)号：US20020198711A1

公开(公告)日：2002-12-26

申请号：US09882744

申请日：2001-06-15

发明人： Yigal Brandman

IPC分类号： G10L015/00 ,  G10L017/00

CPC分类号： G10L15/02 ,  G10L19/0204

摘要： The present invention provides a speech feature extraction system suitable for use in a speech recognition system or other voice processing system that extracts features related to the frequency and amplitude characteristics of an input speech signal using a plurality or complex band pass filters and processing the outputs of adjacent bandpass filters.

摘要翻译： 本发明提供一种适用于语音识别系统或其他语音处理系统的语音特征提取系统，其使用多个或复数带通滤波器提取与输入语音信号的频率和幅度特性相关的特征，并处理 相邻带通滤波器。

公开(公告)号：US20020184023A1

公开(公告)日：2002-12-05

申请号：US09870202

申请日：2001-05-30

发明人： Senis Busayapongchai ,  Pichet Chintrakulchai

IPC分类号： G10L015/00

CPC分类号： H04M3/4936 ,  G10L2015/228

摘要： An interactive speech-activated information retrieval application for use in automated telephone systems includes a control manager that interfaces between the caller's speech input and applications and enables several applications to be open at the same time. The control manager continually monitors for control words, enabling the user to switch between applications at will. When a user switches to another application, the control manager suspends the first application and stores its context, enabling the user to later return to the application at the point where the application was previously suspended.

摘要翻译： 在自动电话系统中使用的交互式语音激活信息检索应用程序包括控制管理器，该控制管理器在呼叫者的语音输入和应用之间进行接口，并且能够同时打开多个应用程序。 控制管理器持续监控控制字，使用户可以随意切换应用程序。 当用户切换到另一个应用程序时，控制管理器暂停第一个应用程序并存储其上下文，使得用户稍后可以在应用程序先前暂停的时候返回到应用程序。

公开(公告)号：US20020173960A1

公开(公告)日：2002-11-21

申请号：US10044464

申请日：2002-01-10

申请人： International Business Machines Corporation

发明人： Steven I. Ross ,  Jeffrey G. MacAllister ,  Julie F. Alweis

IPC分类号： G10L015/18 ,  G10L015/00

CPC分类号： G10L15/1822 ,  G10L15/22 ,  G10L2015/228

摘要： A conversation manager processes spoken utterances from a user of a computer, and develops responses to the spoken utterances. The conversation manager includes a reasoning facility and a language generation module. Each response has a domain model associated with it. The domain model includes an ontology (i.e., world view for the relevant domain of the spoken utterances and responses), lexicon, and syntax definitions. The language generation module receives a response in the form of a formal belief structure from other components of the conversation manager. The reasoning facility selects a syntax template to use in generating a response output from the formal belief structure. The language generation module produces the response output based on the formal structure, the selected syntax template, and the domain model.

摘要翻译： 会话管理员处理来自计算机的用户的讲话，并且发出对讲话语的响应。 会话管理器包括推理设备和语言生成模块。 每个响应都有与之相关联的域模型。 领域模型包括一个本体论（即用于说出话语和响应的相关领域的世界观），词典和语法定义。 语言生成模块以对话管理器的其他组件的正式信念结构的形式接收响应。 推理设施选择用于从形式信念结构生成响应输出的语法模板。 语言生成模块基于形式结构，所选语法模板和域模型生成响应输出。

公开(公告)号：US20020165715A1

公开(公告)日：2002-11-07

申请号：US10020895

申请日：2001-12-19

发明人： Soren Riis ,  Konstantinos Koumpis

IPC分类号： G10L015/00

CPC分类号： G10L15/08

摘要： A speech recognition system uses a phoneme counter to determine the length of a word to be recognized. The result is used to split a lexicon into one or more sub-lexicons containing only words which have the same or similar length to that of the word to be recognized, so restricting the search space significantly. In another aspect, a phoneme counter is used to estimate the number of phonemes in a word so that a transition bias can be calculated. This bias is applied to the transition probabilities between phoneme models in an HNN based recognizer to improve recognition performance for relatively short or long words.

摘要翻译： 语音识别系统使用音素计数器来确定要识别的单词的长度。 结果用于将词典分割成一个或多个仅包含与要识别的词长度相同或相似的字词的子词典，从而显着限制搜索空间。 在另一方面，使用音素计数器来估计单词中的音素的数量，从而可以计算出转变偏差。 该偏差被应用于基于HNN的识别器中的音素模型之间的转换概率，以提高相对较短或长的字的识别性能。

公开(公告)号：US20020156629A1

公开(公告)日：2002-10-24

申请号：US10178807

申请日：2002-06-24

申请人： International Business Machines Corporation

发明人： Martin D. Carberry ,  Mark E. Epstein ,  Glenn T. Puchtel ,  Susan M. Wingate

IPC分类号： G10L015/00

CPC分类号： H04M3/4936 ,  G10L15/1822 ,  Y10S707/99934

摘要： A system and method for servicing natural language requests with a plurality of remote host systems. The system utilizes a computer program that comprises: (1) an input system for inputting an NL command; (2) a translation system that extracts a request from the NL command and stores the request in a host-independent format; and (3) a routing system for servicing the request, wherein the routing system comprises a mechanism for selecting a host, for converting the request into a host dependent directive, and for forwarding the directive to the selected host. The system may further include a voice recognition system, a local data source for servicing the NL command, templates for converting the request into the host dependent directive, a heuristic for selecting the host, and an output system for obtaining and outputting the response. The invention further comprises a context mechanism to interpret natural language instructions, wherein the context mechanism comprises: (1) a context database for storing sets of command elements and sets of response elements; (2) a context requirement mechanism that determines if a current NL command comprised of a current set of command elements is ambiguous; (3) a context retrieving mechanism that retrieves a previous set of response and/or command elements from the context database; and (4) a disambiguation mechanism that uses the retrieved set of response and/or command elements to disambiguate the current set of command elements.

摘要翻译： 一种用于利用多个远程主机系统来服务自然语言请求的系统和方法。 该系统利用计算机程序，其包括：（1）用于输入NL命令的输入系统; （2）翻译系统，其从NL命令中提取请求并将请求存储在与主机无关的格式中; 以及（3）用于对所述请求进行服务的路由系统，其中所述路由系统包括用于选择主机的机制，用于将所述请求转换为主机相关指令，以及将所述指令转发到所选主机。 该系统还可以包括语音识别系统，用于服务NL命令的本地数据源，用于将请求转换为主机相关指令的模板，用于选择主机的启发式，以及用于获得和输出响应的输出系统。 本发明还包括用于解释自然语言指令的上下文机制，其中所述上下文机制包括：（1）用于存储命令元素集合和响应元素集合的上下文数据库; （2）上下文需求机制，其确定由当前命令元素组构成的当前NL命令是否是不明确的; （3）上下文检索机制，其从所述上下文数据库中检索先前的一组响应和/或命令元素; 和（4）消歧机制，其使用所检索的一组响应和/或命令元素来消除当前命令元素集合的歧义。

公开(公告)号：US20020138261A1

公开(公告)日：2002-09-26

申请号：US09814576

申请日：2001-03-22

发明人： Daniel Ziegelmiller

IPC分类号： G10L015/00

CPC分类号： G10L15/26

摘要： The present invention provides a method to automate the validation of dynamic data presented over telecommunications paths. The invention utilizes continuous speaker-independent speech recognition together with a process known generally as natural language recognition to reduce dynamic utterances to machine encoded text without requiring a prior training phase. Further, when configured by the end user to do so, the test system will convert common examples of dynamic speech, such as numbers, dates, times, and currency utterances into their usual textual representation. This eliminates the limitation that all tested utterances need to be known by the test system in advance of the test. By converting the dynamic utterances to machine encoded text, the invention facilitates automated validation of the data so converted, by allowing its use as input into an automated system which can independently access an validate the data.

摘要翻译： 本发明提供了一种使通过电信路径呈现的动态数据的验证自动化的方法。 本发明利用连续的说话者独立的语音识别以及通常被称为自然语言识别的过程，以减少对编码文本的动态发声，而不需要先前的训练阶段。 此外，当由最终用户配置这样做时，测试系统将将动态语音的常见示例（例如数字，日期，时间和货币话语）转换成其通常的文本表示。 这消除了在测试之前测试系统需要知道所有测试语音的限制。 通过将动态话语转换为机器编码文本，本发明通过允许将其作为输入到自动化系统中来进行自动验证，从而可以独立访问验证数据。

公开(公告)号：US20020133338A1

公开(公告)日：2002-09-19

申请号：US10087980

申请日：2002-03-05

申请人： Nissan Motor Co., Ltd.

发明人： Takeshi Ono ,  Okihiko Nakayama

IPC分类号： G10L015/00

CPC分类号： G10L15/22 ,  G10L15/08

摘要： A voice recognition device is provided to improve a recognition rate for objective recognition terms on display. The device includes a voice pickup unit 5 for picking user's voices up, a storing unit for storing a plurality of objective recognition terms, a display unit la for displaying a designated number of objective recognition terms in the objective recognition terms stored in the storing unit and a voice recognition unit 2. The voice recognition unit 2 has a weighting section for weighting the objective recognition terms on display larger than the other objective recognition terms which are not on display, and a calculating section for calculating respective degrees of agreement between the objective recognition terms after weighting and the user's voices picked up by the unit 5. Based on this calculating result of the degrees of agreement, the voice recognition device does recognize the user's voices inputted.

摘要翻译： 提供语音识别装置，以提高显示器上的客观识别项的识别率。 该设备包括用于拾取用户声音的语音拾取单元5，用于存储多个客观识别项的存储单元，用于在存储在存储单元中的目标识别项中显示指定数量的客观识别项的显示单元1a，以及 语音识别单元2.语音识别单元2具有加权部分，用于对显示时的目标识别项进行加权，比不显示的其他目标识别项更大;以及计算部分，用于计算目标识别之间的一致性 加权后的项目以及由单元5拾取的用户的声音。基于协商程度的计算结果，语音识别装置确认识别输入的用户的声音。

公开(公告)号：US20020123891A1

公开(公告)日：2002-09-05

申请号：US09798655

申请日：2001-03-01

申请人： International Business Machines Corporation

发明人： Mark E. Epstein

IPC分类号： G10L015/00

CPC分类号： G10L15/197 ,  G10L15/183

摘要： The invention disclosed herein concerns a method of converting speech to text using a hierarchy of contextual models. The hierarchy of contextual models can be statistically smoothed into a language model. The method can include processing text with a plurality of contextual models. Each one of the plurality of contextual models can correspond to a node in a hierarchy of the plurality of contextual models. Also included can be identifying at least one of the contextual models relating to the text and processing subsequent user spoken utterances with the identified at least one contextual model.

摘要翻译： 本文公开的发明涉及使用上下文模型的层次将语音转换为文本的方法。 语境模型的层次结构可以被统计学平滑化为语言模型。 该方法可以包括用多个上下文模型处理文本。 多个上下文模型中的每一个可以对应于多个上下文模型的层次结构中的节点。 还包括可以识别与文本相关的上下文模型中的至少一个以及使用所识别的至少一个上下文模型来处理随后的用户口语话语。

公开(公告)号：US20020120449A1

公开(公告)日：2002-08-29

申请号：US09796301

申请日：2001-02-28

发明人： Edward O. Clapper

IPC分类号： G10L015/28 ,  G10L015/00

CPC分类号： G10L15/02 ,  G10L15/24

摘要： A characteristic of one or more human resonating cavities may be utilized to provide information for speech recognition, independent from the actual sounds produced. In one embodiment, information about the changing shape of the human oral cavity may provide information useful in determining the nature of a person's vocalizations for speech recognition purposes.

摘要翻译： 可以使用一个或多个人谐振腔的特征来提供用于语音识别的信息，而与所产生的实际声音无关。 在一个实施例中，关于人口腔形状变化的信息可以提供用于确定用于语音识别目的的人的发声的性质的信息。