专利快速检索-快速检索全球专利，免费商用专利数据库-IPRDB

1. 发明申请

US20060184360A1 Adaptive multi-pass speech recognition system 有权
公开(公告)号：US20060184360A1
公开(公告)日：2006-08-17
申请号：US11398402
申请日：2006-04-04
申请人： Hy Murveit , Ashvin Kannan , Ben Shahshahani , Chris Leggetter , Katherine Knill
发明人： Hy Murveit , Ashvin Kannan , Ben Shahshahani , Chris Leggetter , Katherine Knill
IPC分类号： G10L15/00
CPC分类号： G10L15/08 , G10L2015/085
摘要： Method and apparatus for multi-pass speech recognition. An input device receives spoken input. A processor performs a first pass speech recognition technique on the spoken input and forms first pass results. The first pass results include a number of alternative speech expressions, each having an assigned score related to the certainty that the corresponding expression correctly matches the spoken input. The processor selectively performs a second pass speech recognition technique on the spoken input according to the first pass results. Preferably, the second pass attempts to correctly match the spoken input to only those expressions which were identified during the first pass. Otherwise, if one of the expressions identified by the first pass is assigned a score higher than a predetermined threshold (e.g., 95%), the second pass is not performed. Because the second pass is performed only when necessary, the invention recognizes speech with a faster average speed for a given accuracy in comparison to prior systems. Alternately, the first pass results identify a characteristic of the spoken input. The characteristic can be the gender of the speaker or a type of telephone the speaker is calling from. In which case, the second pass speech recognition technique is selected from a plurality of speech recognition techniques according to the characteristic identified by the first pass. Because the selected second pass technique is specific to the characteristic of the spoken input, the second pass technique can perform speech recognition faster for a given accuracy than a technique which is not specific.

2. 发明授权

US07555430B2 Selective multi-pass speech recognition system and method 有权
标题翻译：选择性多遍语音识别系统及方法
公开(公告)号：US07555430B2
公开(公告)日：2009-06-30
申请号：US11398402
申请日：2006-04-04
申请人： Hy Murveit , Ashvin Kannan , Ben Shahshahani , Chris Leggetter , Katherine Knill
发明人： Hy Murveit , Ashvin Kannan , Ben Shahshahani , Chris Leggetter , Katherine Knill
IPC分类号： G10L11/06
CPC分类号： G10L15/08 , G10L2015/085
摘要： Method and apparatus for multi-pass speech recognition. An input device receives spoken input. A processor performs a first pass speech recognition technique on the spoken input and forms first pass results. The first pass results include a number of alternative speech expressions, each having an assigned score related to the certainty that the corresponding expression correctly matches the spoken input. The processor selectively performs a second pass speech recognition technique on the spoken input according to the first pass results. Preferably, the second pass attempts to correctly match the spoken input to only those expressions which were identified during the first pass. Otherwise, if one of the expressions identified by the first pass is assigned a score higher than a predetermined threshold (e.g., 95%), the second pass is not performed. Because the second pass is performed only when necessary, the invention recognizes speech with a faster average speed for a given accuracy in comparison to prior systems.
摘要翻译：用于多遍语音识别的方法和装置。输入设备接收口头输入。处理器在口头输入端执行第一遍语音识别技术，并形成第一遍的结果。第一遍结果包括多个替代语言表达，每个具有与相应表达正确匹配口语输入的确定性相关的分配分数。处理器根据第一遍的结果选择性地在口头输入上执行第二遍语音识别技术。优选地，第二遍尝试将口头输入正确地匹配到在第一遍期间被识别的那些表达。否则，如果由第一遍识别的表达式中的一个被分配高于预定阈值（例如，95％）的分数，则不执行第二遍。因为仅在必要时执行第二遍，所以与现有系统相比，本发明以给定精度更快的平均速度识别语音。

3. 发明申请

US20060178879A1 Adaptive multi-pass speech recognition system 有权
标题翻译：自适应多遍语音识别系统
公开(公告)号：US20060178879A1
公开(公告)日：2006-08-10
申请号：US11398327
申请日：2006-04-04
申请人： Hy Murveit , Ashvin Kannan , Ben Shahshahani , Chris Leggetter , Katherine Knill
发明人： Hy Murveit , Ashvin Kannan , Ben Shahshahani , Chris Leggetter , Katherine Knill
IPC分类号： G10L15/00
CPC分类号： G10L15/08 , G10L2015/085
摘要： Method and apparatus for multi-pass speech recognition. An input device receives spoken input. A processor performs a first pass speech recognition technique on the spoken input and forms first pass results. The first pass results include a number of alternative speech expressions, each having an assigned score related to the certainty that the corresponding expression correctly matches the spoken input. The processor selectively performs a second pass speech recognition technique on the spoken input according to the first pass results. Preferably, the second pass attempts to correctly match the spoken input to only those expressions which were identified during the first pass. Otherwise, if one of the expressions identified by the first pass is assigned a score higher than a predetermined threshold (e.g., 95%), the second pass is not performed. Because the second pass is performed only when necessary, the invention recognizes speech with a faster average speed for a given accuracy in comparison to prior systems. Alternately, the first pass results identify a characteristic of the spoken input. The characteristic can be the gender of the speaker or a type of telephone the speaker is calling from. In which case, the second pass speech recognition technique is selected from a plurality of speech recognition techniques according to the characteristic identified by the first pass. Because the selected second pass technique is specific to the characteristic of the spoken input, the second pass technique can perform speech recognition faster for a given accuracy than a technique which is not specific.
摘要翻译：用于多遍语音识别的方法和装置。输入设备接收口头输入。处理器在口头输入端执行第一遍语音识别技术，并形成第一遍的结果。第一遍结果包括多个替代语言表达，每个具有与相应表达正确匹配口语输入的确定性相关的分配分数。处理器根据第一遍的结果选择性地在口头输入上执行第二遍语音识别技术。优选地，第二遍尝试将口头输入正确地匹配到在第一遍期间被识别的那些表达。否则，如果由第一遍识别的表达式中的一个被分配高于预定阈值（例如，95％）的分数，则不执行第二遍。因为仅在必要时执行第二遍，所以与现有系统相比，本发明以给定精度更快的平均速度识别语音。或者，第一遍结果识别口语输入的特征。特征可以是扬声器的性别或扬声器所呼叫的电话类型。在这种情况下，根据由第一遍识别的特征，从多个语音识别技术中选择第二遍语音识别技术。由于所选择的第二传递技术特定于口头输入的特征，所以与非特定的技术相比，第二遍技术可以比给定的精度更快地执行语音识别。

4. 发明授权

US07401017B2 Adaptive multi-pass speech recognition system 有权
标题翻译：自适应多遍语音识别系统
公开(公告)号：US07401017B2
公开(公告)日：2008-07-15
申请号：US11398327
申请日：2006-04-04
申请人： Hy Murveit , Ashvin Kannan , Ben Shahshahani , Chris Leggetter , Katherine Knill
发明人： Hy Murveit , Ashvin Kannan , Ben Shahshahani , Chris Leggetter , Katherine Knill
IPC分类号： G10L21/00
CPC分类号： G10L15/08 , G10L2015/085
摘要： Method and apparatus for multi-pass speech recognition. An input device receives spoken input. A processor performs a first pass speech recognition technique on the spoken input and forms first pass results. The first pass results include a number of alternative speech expressions, each having an assigned score related to the certainty that the corresponding expression correctly matches the spoken input. The processor selectively performs a second pass speech recognition technique on the spoken input according to the first pass results. Preferably, the second pass attempts to correctly match the spoken input to only those expressions which were identified during the first pass. Otherwise, if one of the expressions identified by the first pass is assigned a score higher than a predetermined threshold (e.g., 95%), the second pass is not performed. Because the second pass is performed only when necessary, the invention recognizes speech with a faster average speed for a given accuracy in comparison to prior systems. Alternately, the first pass results identify a characteristic of the spoken input. The characteristic can be the gender of the speaker or a type of telephone the speaker is calling from. In which case, the second pass speech recognition technique is selected from a plurality of speech recognition techniques according to the characteristic identified by the first pass. Because the selected second pass technique is specific to the characteristic of the spoken input, the second pass technique can perform speech recognition faster for a given accuracy than a technique which is not specific.
摘要翻译：用于多遍语音识别的方法和装置。输入设备接收口头输入。处理器在口头输入端执行第一遍语音识别技术，并形成第一遍的结果。第一遍结果包括多个替代语言表达，每个具有与相应表达正确匹配口语输入的确定性相关的分配分数。处理器根据第一遍的结果选择性地在口头输入上执行第二遍语音识别技术。优选地，第二遍尝试将口头输入正确地匹配到在第一遍期间被识别的那些表达。否则，如果由第一遍识别的表达式中的一个被分配高于预定阈值（例如，95％）的分数，则不执行第二遍。因为仅在必要时执行第二遍，所以与现有系统相比，本发明以给定精度更快的平均速度识别语音。或者，第一遍结果确定了口头输入的特征。特征可以是扬声器的性别或扬声器所呼叫的电话类型。在这种情况下，根据由第一遍识别的特征，从多个语音识别技术中选择第二遍语音识别技术。由于所选择的第二传递技术特定于口头输入的特征，所以与非特定的技术相比，第二遍技术可以比给定的精度更快地执行语音识别。

5. 发明授权

US07058573B1 Speech recognition system to selectively utilize different speech recognition techniques over multiple speech recognition passes 有权
标题翻译：语音识别系统，用于在多个语音识别传递中有选择地利用不同的语音识别技术
公开(公告)号：US07058573B1
公开(公告)日：2006-06-06
申请号：US09294847
申请日：1999-04-20
申请人： Hy Murveit , Ashvin Kannan , Ben Shahshahani , Chris Leggetter , Katherine Knill
发明人： Hy Murveit , Ashvin Kannan , Ben Shahshahani , Chris Leggetter , Katherine Knill
IPC分类号： G10L19/02
CPC分类号： G10L15/08 , G10L2015/085
摘要： Method and apparatus for multi-pass speech recognition. An input device receives spoken input. A processor performs a first pass speech recognition technique on the spoken input and forms first pass results. The first pass results include a number of alternative speech expressions, each having an assigned score related to the certainty that the corresponding expression correctly matches the spoken input. The processor selectively performs a second pass speech recognition technique on the spoken input according to the first pass results. Preferably, the second pass attempts to correctly match the spoken input to only those expressions which were identified during the first pass. Otherwise, if one of the expressions identified by the first pass is assigned a score higher than a predetermined threshold (e.g., 95%), the second pass is not performed. Because the second pass is performed only when necessary, the invention recognizes speech with a faster average speed for a given accuracy in comparison to prior systems. Alternately, the first pass results identify a characteristic of the spoken input. The characteristic can be the gender of the speaker or a type of telephone the speaker is calling from. In which case, the second pass speech recognition technique is selected from a plurality of speech recognition techniques according to the characteristic identified by the first pass. Because the selected second pass technique is specific to the characteristic of the spoken input, the second pass technique can perform speech recognition faster for a given accuracy than a technique which is not specific.
摘要翻译：用于多遍语音识别的方法和装置。输入设备接收口头输入。处理器在口头输入端执行第一遍语音识别技术，并形成第一遍的结果。第一遍结果包括多个替代语言表达，每个具有与相应表达正确匹配口语输入的确定性相关的分配分数。处理器根据第一遍的结果选择性地在口头输入上执行第二遍语音识别技术。优选地，第二遍尝试将口头输入正确地匹配到在第一遍期间被识别的那些表达。否则，如果由第一遍识别的表达式中的一个被分配高于预定阈值（例如，95％）的分数，则不执行第二遍。因为仅在必要时执行第二遍，所以与现有系统相比，本发明以给定精度更快的平均速度识别语音。或者，第一遍结果识别口语输入的特征。特征可以是扬声器的性别或扬声器所呼叫的电话类型。在这种情况下，根据由第一遍识别的特征，从多个语音识别技术中选择第二遍语音识别技术。由于所选择的第二传递技术特定于口头输入的特征，所以与非特定的技术相比，第二遍技术可以比给定的精度更快地执行语音识别。

6. 发明授权

US06766295B1 Adaptation of a speech recognition system across multiple remote sessions with a speaker 有权
标题翻译：通过扬声器在多个远程会话之间适应语音识别系统
公开(公告)号：US06766295B1
公开(公告)日：2004-07-20
申请号：US09309211
申请日：1999-05-10
申请人： Hy Murveit , Ashvin Kannan
发明人： Hy Murveit , Ashvin Kannan
IPC分类号： G10L1506
CPC分类号： G10L15/30 , G10L2015/0638
摘要： A technique for adaptation of a speech recognizing system across multiple remote communication sessions with a speaker. The speaker can be a telephone caller. An acoustic model is utilized for recognizing the speaker's speech. Upon initiation of a first remote session with the speaker, the acoustic model is speaker-independent. During the first session, the speaker is uniquely identified and speech samples are obtained from the speaker. In the preferred embodiment, the samples are obtained without requiring the speaker to engage in a training session. The acoustic model is then modified based upon the samples thereby forming a modified model. The model can be modified during the session or after the session is terminated. Upon termination of the session, the modified model is then stored in association with an identification of the speaker. During a subsequent remote session, the speaker is identified and, then, the modified acoustic model is utilized to recognize the speaker's speech. Additional speech samples are obtained during the subsequent session and, then, utilized to further modify the acoustic model. In this manner, an acoustic model utilized for recognizing the speech of a particular speaker is cumulatively modified according to speech samples obtained during multiple sessions with the speaker. As a result, the accuracy of the speech recognizing system improves for the speaker even when the speaker only engages in relatively short remote sessions.
摘要翻译：一种用于通过与扬声器的多个远程通信会话来适应语音识别系统的技术。扬声器可以是电话呼叫者。声学模型被用于识别说话人的言语。在与扬声器开始第一次远程会话时，声学模型是与扬声器无关的。在第一次会话期间，扬声器被唯一地识别，并且从扬声器获得语音样本。在优选实施例中，在不要求说话者参与训练课程的情况下获得样本。然后基于样品修改声学模型，从而形成修改后的模型。该模型可以在会话期间或会话终止后进行修改。在会话结束时，然后将修改的模型与扬声器的标识相关联地存储。在随后的远程会话期间，识别扬声器，然后利用修改的声学模型来识别扬声器的语音。在随后的会话中获得附加语音样本，然后用于进一步修改声学模型。以这种方式，用于识别特定扬声器的语音的声学模型根据在与扬声器的多个会话期间获得的语音样本而被累积地修改。结果，即使扬声器只进行相对较短的远程会话，语音识别系统的精度也提高了。

7. 发明授权

US06728677B1 Method and system for dynamically improving performance of speech recognition or other speech processing systems 有权
标题翻译：用于动态提高语音识别或其他语音处理系统性能的方法和系统
公开(公告)号：US06728677B1
公开(公告)日：2004-04-27
申请号：US09773996
申请日：2001-01-31
申请人： Ashvin Kannan , Hy Murveit , Christopher Leggetter , Michael Schuster
发明人： Ashvin Kannan , Hy Murveit , Christopher Leggetter , Michael Schuster
IPC分类号： G10L2100
CPC分类号： G10L15/285
摘要： The present invention introduces a system and method for dynamically improving speech recognition in a speech recognition or other speech processing system. The method comprises dynamically adjusting the system, which comprises estimating the utilization of resources in the system; and improving the performance of the system according to the availability of resources.
摘要翻译：本发明引入了用于在语音识别或其他语音处理系统中动态改善语音识别的系统和方法。该方法包括动态调整系统，包括估计系统中资源的利用率; 并根据资源的可用性提高系统的性能。

8. 发明申请

US20120030011A1 Systems and Methods for Estimating a Conversion Rate for a Digital Advertisement Based on Dwell Times Associated with the Digital Advertisement 审中-公开
标题翻译：用于估计与数字广告相关的停留时间的数字广告的转换速率的系统和方法
公开(公告)号：US20120030011A1
公开(公告)日：2012-02-02
申请号：US12847530
申请日：2010-07-30
申请人： Benjamin Rey , Ashvin Kannan
发明人： Benjamin Rey , Ashvin Kannan
IPC分类号： G06Q30/00
CPC分类号： G06Q30/0242 , G06Q30/0244 , G06Q30/0272
摘要： The present disclosure is directed to systems and methods for estimating a conversion rate associated with a digital ad. Generally, an ad provider and/or an ad campaign management system detects an action of a user associated with a digital ad. The ad provider and/or ad campaign management system detects a subsequent action of the user after detecting the action of the user associated with the digital ad and determines an amount of time between the action of the user associated with the digital ad and the subsequent action of the user. The ad provider and/or ad campaign management system then estimates a conversion rate associated with the digital ad based on the determined amount of time. In some implementations, a model is utilized to estimate a conversion rate associated with a digital ad based on dwell times even when the ad provider and/or ad campaign management system has little or no conversion data associated with the digital ad.
摘要翻译：本公开涉及用于估计与数字广告相关联的转换率的系统和方法。通常，广告提供商和/或广告活动管理系统检测与数字广告相关联的用户的动作。广告提供商和/或广告活动管理系统在检测到与数字广告相关联的用户的动作之后检测用户的后续动作，并且确定与数字广告相关联的用户的动作与后续动作之间的时间量的用户。然后，广告提供商和/或广告系列管理系统基于所确定的时间量来估计与数字广告相关联的转化率。在一些实施中，即使当广告提供商和/或广告活动管理系统几乎没有或没有与数字广告相关联的转换数据时，也使用模型来估计与数字广告相关联的转换率。

9. 发明申请

US20110040616A1 SPONSORED SEARCH BID ADJUSTMENT BASED ON PREDICTED CONVERSION RATES 审中-公开
标题翻译：基于预测转换率的辅助搜索引擎调整
公开(公告)号：US20110040616A1
公开(公告)日：2011-02-17
申请号：US12541453
申请日：2009-08-14
申请人： Ashvin Kannan , Benjamin Rey
发明人： Ashvin Kannan , Benjamin Rey
IPC分类号： G06Q30/00 , G06F15/18 , G06N5/02 , G06F17/30
CPC分类号： G06Q30/02 , G06Q30/0246 , G06Q30/0275
摘要： Methods and systems are provided for adjusting an advertiser bid, in a sponsored search auction, in connection with one or more advertisements to be served in connection with keyword queries, the bid being associated with one or more keyword phrases. The bid is adjusted based on a predicted conversion rate associated with an advertisement served in connection with a match between a keyword query and the one or more keyword phrases. A bid may be decreased for a match with a lower predicted conversion rate than a comparison predicted conversion rate such as a normalized average predicted conversion rate.
摘要翻译：提供了方法和系统，用于在赞助的搜索拍卖中，结合与关键词查询相关联的一个或多个广告来调整广告主投标，该投标与一个或多个关键词短语相关联。基于与关键词查询与一个或多个关键词短语之间的匹配相关联的广告相关联的预测转换率来调整出价。对于具有比比较预测转换率（例如归一化的平均预测转换率）更低的预测转换速率的匹配，可以降低出价。

10. 发明申请

US20070112840A1 System and method for generating functions to predict the clickability of advertisements 有权
标题翻译：用于生成功能以预测广告的可点击性的系统和方法
公开(公告)号：US20070112840A1
公开(公告)日：2007-05-17
申请号：US11479186
申请日：2006-06-29
申请人： Chad Carson , Ashvin Kannan , Erick Cantu-Paz , Rukmini Iyer , Pero Subasic , Christopher LuVogt , Christopher Leggetter , Jan Pedersen , David Ku
发明人： Chad Carson , Ashvin Kannan , Erick Cantu-Paz , Rukmini Iyer , Pero Subasic , Christopher LuVogt , Christopher Leggetter , Jan Pedersen , David Ku
IPC分类号： G06F7/00
CPC分类号： G06Q10/04 , G06Q30/02 , G06Q30/0202 , G06Q30/0246 , Y10S707/99933 , Y10S707/99943
摘要： The present invention is directed towards systems and methods for predicting a frequency with which an advertisement displayed in response to a query will be selected. The method of the present invention comprises receiving analytics data associated with a display of one or more advertisements in response to one or more queries. One or more features associated with the one or more advertisements displayed in response to the one or more queries are identified. One or more functions are generated for predicting a frequency with which a given advertisement displayed in response to a query will be selected using the analytics data and features associated with the one or more advertisements displayed in response to the one or more queries.
摘要翻译：本发明涉及用于预测响应于查询显示的广告的频率的系统和方法。本发明的方法包括响应于一个或多个查询而接收与一个或多个广告的显示相关联的分析数据。识别与响应于一个或多个查询而显示的一个或多个广告相关联的一个或多个特征。生成一个或多个功能，用于预测使用响应于一个或多个查询显示的一个或多个广告相关联的分析数据和特征来选择响应于查询而显示的给定广告的频率。

你已经成功收藏专利！

检索式保存成功!

IPRDB

热门服务

关于我们

友情链接

联系方式