公开(公告)号：US5365574A

公开(公告)日：1994-11-15

申请号：US982186

申请日：1992-11-25

申请人： Alan K. Hunt ,  Thomas B. Schalk

发明人： Alan K. Hunt ,  Thomas B. Schalk

IPC分类号： G07C9/00 ,  G10L17/00 ,  H04M3/38 ,  G10L9/08 ,  H04M1/66

CPC分类号： G07C9/00142 ,  G06F21/32 ,  G06F21/40 ,  G06F21/83 ,  G07C9/00158 ,  G10L17/00 ,  G10L17/22 ,  G10L17/24 ,  H04M3/382 ,  G06F2221/2105 ,  G06F2221/2113 ,  G06F2221/2117 ,  H04M2201/40

摘要： A method for enabling a caller to obtain access to one or more services via a telephone network by speaking a password is described. The method begins by establishing at least one predetermined threshold value for a speaker verification signal. For each spoken character of the password, the method generates a set of parameters using a voice verification feature transformation. After entry and recognition of the password, the sets of parameters are used to generate a speaker verification signal for the password. Upon the occurrence of a predetermined call condition, the predetermined threshold value is adjusted to increase the level of security in the verification algorithm. If the signal verification signal for the password has a predetermined relationship with respect to the adjusted threshold value, the caller's identity is accepted. If not, the caller may be asked to answer certain personal questions before his or her identity is accepted.

摘要翻译： 描述了一种使呼叫者能够通过说出口令通过电话网络获得对一个或多个服务的访问的方法。 该方法通过为扬声器验证信号建立至少一个预定阈值开始。 对于密码的每个口头字符，该方法使用语音验证特征变换生成一组参数。 在输入和识别密码后，使用这些参数组来生成用于密码的扬声器验证信号。 在出现预定呼叫条件时，调整预定阈值以增加验证算法中的安全级别。 如果用于密码的信号验证信号相对于经调整的阈值具有预定的关系，则接受呼叫者的身份。 如果没有，呼叫者可能会被要求在他或她的身份被接受之前回答某些个人问题。

公开(公告)号：US5297194A

公开(公告)日：1994-03-22

申请号：US901742

申请日：1992-06-22

申请人： Alan K. Hunt ,  Thomas B. Schalk

发明人： Alan K. Hunt ,  Thomas B. Schalk

IPC分类号： G07C9/00 ,  G10L17/00 ,  H04M3/38 ,  G10L9/08 ,  H04M1/66

CPC分类号： G06F21/32 ,  G06F21/40 ,  G06F21/83 ,  G07C9/00142 ,  G07C9/00158 ,  G10L17/00 ,  G10L17/22 ,  G10L17/24 ,  H04M3/382 ,  G06F2221/2105 ,  G06F2221/2113 ,  G06F2221/2117 ,  H04M2201/40

摘要： The present invention describes a system and method for enabling a caller to obtain access to services via a telephone network by entering a spoken password having a plurality of digits. Preferably, the method includes the steps of: (1) prompting the caller to speak the password beginning with a first digit and ending with a last digit thereof, (2) recognizing each spoken digit of the password using a speaker-independent voice recognition algorithm, (3) following entry of the last digit of the password, determining whether the password is valid, and (4) if the password is valid, verifying the caller's identity using a voice verification algorithm.

摘要翻译： 本发明描述了一种用于使呼叫者能够通过输入具有多个数字的口语密码通过电话网获取服务的系统和方法。 优选地，该方法包括以下步骤：（1）提示呼叫者以第一个数字开始并以其最后一个数字结尾来说出口令，（2）使用与说话者无关的语音识别算法来识别密码的每个口头数字 ，（3）在输入密码的最后一位数字后，确定密码是否有效，以及（4）如果密码有效，则使用语音验证算法验证主叫的身份。

公开(公告)号：US5303299A

公开(公告)日：1994-04-12

申请号：US901759

申请日：1992-06-22

申请人： Alan K. Hunt ,  Thomas B. Schalk

发明人： Alan K. Hunt ,  Thomas B. Schalk

IPC分类号： G07C9/00 ,  G10L17/00 ,  H04M3/38 ,  G10L9/08 ,  H04M1/66

CPC分类号： G10L17/24 ,  G07C9/00158 ,  H04M3/382 ,  H04M2201/40

摘要： The present invention describes a method for recognizing alphanumeric strings spoken over a telephone network wherein individual character recognition need not be uniformly high in order to achieve high string recognition accuracy. Preferably, the method uses a processing system having a digital processor, an interface to the telephone network, and a database for storing a predetermined set of reference alphanumeric strings. In operation, the system prompts the caller to speak the characters of a string, and characters are recognized using a speaker-independent voice recognition algorithm. The method calculates recognition distances between each spoken input character and the corresponding letter or digit in the same position within each reference alphanumeric string. After each character is spoken, captured and analyzed, each reference string distance is incremented and the process is continued, accumulating distances for each reference string, until the last character is spoken. The reference string with the lowest cumulative distance is then declared to be the recognized string.

摘要翻译： 本发明描述了一种用于识别通过电话网络说出的字母数字串的方法，其中为了实现高的字符串识别精度，单独的字符识别不需要均匀地高。 优选地，该方法使用具有数字处理器，到电话网络的接口的处理系统和用于存储预定集合的参考字母数字串的数据库。 在操作中，系统提示呼叫者说出字符串的字符，并且使用与说话者无关的语音识别算法来识别字符。 该方法计算每个口头输入字符与每个参考字母数字串中相同位置的相应字母或数字之间的识别距离。 在每个字符被说出，捕获和分析之后，每个参考字符串距离递增，并且处理继续，累积每个参考字符串的距离，直到说出最后一个字符。 具有最小累积距离的参考字符串然后被声明为识别的字符串。

公开(公告)号：US5499288A

公开(公告)日：1996-03-12

申请号：US216399

申请日：1994-03-22

申请人： Alan K. Hunt ,  Thomas B. Schalk

发明人： Alan K. Hunt ,  Thomas B. Schalk

IPC分类号： G07C9/00 ,  G10L17/00 ,  H04M3/38 ,  G10L9/08 ,  H04M1/66

CPC分类号： G06F21/32 ,  G06F21/40 ,  G06F21/83 ,  G07C9/00142 ,  G07C9/00158 ,  G10L17/00 ,  G10L17/22 ,  G10L17/24 ,  H04M3/382 ,  G06F2221/2105 ,  G06F2221/2113 ,  G06F2221/2117 ,  H04M2201/40

摘要： The present invention describes a system and method for enabling a caller to obtain access to services via a telephone network by entering a spoken password having a plurality of digits. Preferably, the method includes the steps of: (1) prompting the caller to speak the password beginning with a first digit and ending with a last digit thereof, (2) recognizing each spoken digit of the password using a speaker-independent voice recognition algorithm, (3) following entry of the last digit of the password, determining whether the password is valid, and (4) if the password is valid, verifying the caller's identity using a voice verification algorithm.

摘要翻译： 本发明描述了一种用于使呼叫者能够通过输入具有多个数字的口语密码通过电话网获取服务的系统和方法。 优选地，该方法包括以下步骤：（1）提示呼叫者以第一个数字开始并以其最后一个数字结尾来说出口令，（2）使用与说话者无关的语音识别算法来识别密码的每个口头数字 ，（3）在输入密码的最后一位数字后，确定密码是否有效，以及（4）如果密码有效，则使用语音验证算法验证主叫的身份。

公开(公告)号：US5127043A

公开(公告)日：1992-06-30

申请号：US523486

申请日：1990-05-15

申请人： Alan K. Hunt ,  Thomas B. Schalk

发明人： Alan K. Hunt ,  Thomas B. Schalk

IPC分类号： G07C9/00 ,  G10L17/00 ,  H04M3/38

CPC分类号： G06F21/32 ,  G06F21/40 ,  G06F21/83 ,  G07C9/00142 ,  G07C9/00158 ,  G10L17/00 ,  G10L17/22 ,  G10L17/24 ,  H04M3/382 ,  G06F2221/2105 ,  G06F2221/2113 ,  G06F2221/2117 ,  H04M2201/40

摘要： The present invention describes a system and method for enabling a caller to obtain access to services via a telephone network by entering a spoken password having a plurality of digits. Preferably, the method includes the steps of: (1) prompting the caller to speak the password beginning with a first digit and ending with a last digit thereof, (2) recognizing each spoken digit of the password using a speaker-independent voice recognition algorithm, (3) following entry of the last digit of the password, determining whether the password is valid, and (4) if the password is valid, verifying the caller's identity using a voice verification algorithm.

摘要翻译： 本发明描述了一种用于使呼叫者能够通过输入具有多个数字的口语密码通过电话网获取服务的系统和方法。 优选地，该方法包括以下步骤：（1）提示呼叫者以第一个数字开始并以其最后一个数字结尾来说出口令，（2）使用与说话者无关的语音识别算法来识别密码的每个口头数字 ，（3）在输入密码的最后一位数字后，确定密码是否有效，以及（4）如果密码有效，则使用语音验证算法验证主叫的身份。

公开(公告)号：US5125022A

公开(公告)日：1992-06-23

申请号：US566519

申请日：1990-08-10

申请人： Alan K. Hunt ,  Thomas B. Schalk

发明人： Alan K. Hunt ,  Thomas B. Schalk

IPC分类号： G07C9/00 ,  G10L15/10 ,  G10L15/22 ,  G10L17/00 ,  H04M3/38

CPC分类号： G10L17/24 ,  G07C9/00158 ,  H04M3/382 ,  G10L15/10 ,  G10L2015/0631 ,  H04M2201/40

摘要： The present invention describes a method for recognizing alphanumeric strings spoken over a telephone network wherein individual character recognition need not be uniformly high in order to achieve high string recognition accuracy. Preferably, the method uses a processing system having a digital processor, an interface to the telephone network, and a database for storing a predetermined set of reference alphanumeric strings. In operation, the system prompts the caller to speak each character of a string, beginning with a first character and ending with a last character. Each character is then recognized using a speaker-independent voice recognition algorithm. The method calculates recognition distances between each spoken input character and the corresponding letter or digit in the same position within each reference alphanumeric string. After each character is spoken, captured and analyzed, each reference string distance is incremented and the process is continued, accumulating distances for each reference string, until the last character is spoken. The reference string with the lowest cumulative distance is then declared to be the recognized string.

摘要翻译： 本发明描述了一种用于识别通过电话网络说出的字母数字串的方法，其中为了实现高的字符串识别精度，单独的字符识别不需要均匀地高。 优选地，该方法使用具有数字处理器，到电话网络的接口的处理系统和用于存储预定集合的参考字母数字串的数据库。 在操作中，系统提示呼叫者说出字符串的每个字符，从第一个字符开始，以最后一个字符结束。 然后，使用与扬声器无关的语音识别算法来识别每个字符。 该方法计算每个口头输入字符与每个参考字母数字串中相同位置的相应字母或数字之间的识别距离。 在每个字符被说出，捕获和分析之后，每个参考字符串距离递增，并且处理继续，累积每个参考字符串的距离，直到说出最后一个字符。 具有最小累积距离的参考字符串然后被声明为识别的字符串。

公开(公告)号：US5659597A

公开(公告)日：1997-08-19

申请号：US216009

申请日：1994-03-22

申请人： Bernard F. Bareis ,  Peter J. Foster ,  Thomas B. Schalk

发明人： Bernard F. Bareis ,  Peter J. Foster ,  Thomas B. Schalk

IPC分类号： H04B7/185 ,  H04M3/42 ,  H04M3/493 ,  H04M11/10 ,  H04W4/00 ,  H04W4/18 ,  H04W88/18 ,  H04M1/66 ,  G10L9/08 ,  H04Q7/22

CPC分类号： H04W4/18 ,  H04B7/18567 ,  H04M11/10 ,  H04M3/42204 ,  H04M3/4936 ,  H04W4/00 ,  H04M2201/40 ,  H04M2201/405 ,  H04W88/18

摘要： An advanced telecommunications system is provided for the recognizing of spoken commands over a cellular telephone, satellite telephone, or personal communications network. In the cellular application, for example, a Speech Recognition System interconnects either internally with or as an external peripheral to a cellular telecommunications switch. The Speech Recognition System includes an administrative subsystem, a call processing subsystem, a speaker-dependent recognition subsystem, a speaker-independent recognition subsystem, and a data storage subsystem. The Speech Recognition System also allows for increased efficiency in the cellular telephone network by integrating with the switch or switches as a shared resource. The administrative subsystem of the Speech Recognition System is used to keep statistical logs of pertinent call information. Pre-recorded instructional messages are stored in the memory of the call processing subsystem for instructing a user on his or her progress in using the system. The speaker-independent recognition subsystem allows the user to interact with the system employing non-user specific functions. User specific functions are controlled with the speaker-dependent recognition subsystem. User specific attributes collected by the recognition subsystems are stored in the data storage subsystem.

摘要翻译： 提供了一种先进的电信系统，用于通过蜂窝电话，卫星电话或个人通信网络识别语音命令。 在蜂窝应用中，例如，语音识别系统在蜂窝电信交换机内部或作为外部外围设备互连。 语音识别系统包括管理子系统，呼叫处理子系统，与扬声器相关的识别子系统，与扬声器无关的识别子系统和数据存储子系统。 语音识别系统还允许通过与交换机或交换机作为共享资源集成来提高蜂窝电话网络的效率。 语音识别系统的管理子系统用于保存相关呼叫信息的统计日志。 预先记录的教学消息存储在呼叫处理子系统的存储器中，用于指示用户在使用系统方面的进展。 与扬声器无关的识别子系统允许用户与采用非用户特定功能的系统交互。 用户特定的功能由与扬声器相关的识别子系统控制。 由识别子系统收集的用户特定属性存储在数据存储子系统中。

公开(公告)号：US20090074157A1

公开(公告)日：2009-03-19

申请号：US12326753

申请日：2008-12-02

申请人： Bernard F. Bareis ,  Peter J. Foster ,  Thomas B. Schalk

发明人： Bernard F. Bareis ,  Peter J. Foster ,  Thomas B. Schalk

IPC分类号： H04M1/64

CPC分类号： H04W4/18 ,  H04B7/18567 ,  H04M3/42204 ,  H04M3/4936 ,  H04M11/10 ,  H04M2201/40 ,  H04M2201/405 ,  H04W4/00 ,  H04W88/18

摘要： Voice activated dialing is described for use in a mobile telecommunications system. A voice input is received from a wireless network user. A telephone number to be dialed is determined by using speaker independent speech recognition to interpret a string of spoken digits in the voice input to determine the telephone number, or using speaker dependent speech recognition to interpret a spoken word in the voice input to determine the telephone number. A telephone call is then initiated by dialing the telephone number.

摘要翻译： 语音激活拨号被描述为用于移动电信系统。 从无线网络用户接收语音输入。 要拨打的电话号码通过使用与讲话者无关的语音识别来解释语音输入中的一串口头数字来确定电话号码，或者使用说话者依赖语音识别来解释语音输入中的口语单词以确定电话 数。 然后通过拨打电话号码开始电话呼叫。

公开(公告)号：US4989248A

公开(公告)日：1991-01-29

申请号：US319384

申请日：1989-03-03

申请人： Thomas B. Schalk ,  George R. Doddington

发明人： Thomas B. Schalk ,  George R. Doddington

IPC分类号： G10L15/00 ,  G10L15/12

CPC分类号： G10L15/12

摘要： A cost-effective word recognizer. Each frame of spoken input is compared to a set of reference frames. The comparison is equivalent to embodying the reference frame as an LPC inverse filter, and is preferably done in the autocorrelation domain. To avoid the instability and computational difficulties which can be caused by a high-gain LPC inverse filter, a noise floor is introduced into each reference frame sample. Thus, for each input speech frame, a scalar measures its similarity to each of the vocabulary of reference frames.To achieve connected word recognition based on this similarity measurement, a dynamic programming algorithm is used in which time warping to match a sample to a reference is in effect permitted, and in which matching is performed with unconstrained endpoints. Thus, the word boundary decisions are made on the basis of a local maximum in similarity, and, since no separate word division decision is required, the error which can be introduced by even the best preliminary decision as to word boundaries is avoided.

摘要翻译： 一个具有成本效益的字识别器。 将每个语音输入帧与一组参考帧进行比较。 该比较等效于将参考帧作为LPC逆滤波器，并且优选地在自相关域中完成。 为了避免由高增益LPC反相滤波器引起的不稳定性和计算困难，本底噪声被引入每个参考帧样本。 因此，对于每个输入语音帧，标量测量其与参考帧的每个词汇表的相似度。 为了实现基于这种相似性测量的连续字识别，使用动态规划算法，其中实际上允许将样本与参考匹配的时间扭曲，并且其中使用非约束端点执行匹配。 因此，单词边界决策是基于相似度的局部最大值进行的，并且由于不需要单独的分离决策，所以避免了即使是关于字边界的最佳初步决定也可以引入的误差。

公开(公告)号：US06922669B2

公开(公告)日：2005-07-26

申请号：US09222073

申请日：1998-12-29

申请人： Thomas B. Schalk ,  Roger S. Zimmerman

发明人： Thomas B. Schalk ,  Roger S. Zimmerman

IPC分类号： G06F3/16 ,  G10L15/00 ,  G10L15/18 ,  G10L15/22 ,  G10L15/28 ,  G10L15/10

CPC分类号： G10L15/18 ,  G10L2015/085

摘要： A highly accurate technique for recognizing spoken digit strings is described. A spoken digit string is received and analyzed by a speech recognizer, which generates a list of hypothesized digit strings arranged in ranked order based on a likelihood of matching the spoken digit string. The individual hypothesized strings are then analyzed in order beginning with the hypothesized string having the greatest likelihood of matching the spoken string to determine whether they satisfy a given constraint. The first hypothesized string in the list satisfying the constraint is selected as the recognized string.

摘要翻译： 描述了用于识别口头字符串的高精度技术。 语音数字串由语音识别器接收和分析，语音识别器基于与口语数字串匹配的可能性来生成以排序顺序排列的假设数字串列表。 然后，以具有匹配口语字符串的最大可能性的假设字符串开始，按顺序分析各个假设字符串，以确定它们是否满足给定的约束。 选择满足约束的列表中的第一个假设字符串作为识别的字符串。