-
公开(公告)号：US20090259466A1
公开(公告)日：2009-10-15
申请号：US12423527
申请日：2009-04-14
Applicant: Peter Stubley , Oded Monzon , Alex Gorodetski , Douglas Peters
Inventor： Peter Stubley , Oded Monzon , Alex Gorodetski , Douglas Peters
IPC: G10L15/14
CPC classification number: G10L15/08 , G10L2015/0631
Abstract: Adjusting confidence score thresholds is described for a speech recognition engine. The speech recognition engine is implemented in multiple computer processes functioning in a computer processor, and is characterized by an associated receiver operating characteristic (ROC) curve. A results confirmation process interprets user confirmation of speech recognition results within a given confidence score threshold to create a confirmed portion of the ROC curve for the speech recognition engine. A curve extension process extends the confirmed portion of the ROC curve by extrapolation of unconfirmed speech recognition results beyond the confidence score threshold to generate an extended ROC curve. A threshold adjustment process adjusts the confidence score threshold based on the extended ROC curve to meet target operating constraints for operating the speech recognition engine to perform automatic speech recognition of user speech inputs.
Abstract translation: 为语音识别引擎描述了调整置信度分数阈值。 语音识别引擎在计算机处理器中运行的多个计算机进程中实现，并且其特征在于相关联的接收器操作特性（ROC）曲线。 结果确认过程在给定的置信分数阈值内解释用户对语音识别结果的确认，以创建用于语音识别引擎的ROC曲线的确认部分。 曲线扩展过程通过外推未确认的语音识别结果超出置信分数阈值来扩展ROC曲线的确认部分，以产生扩展的ROC曲线。 阈值调整过程基于扩展的ROC曲线来调整置信分数阈值，以满足用于操作语音识别引擎以执行用户语音输入的自动语音识别的目标操作约束。
-
公开(公告)号：US20050152511A1
公开(公告)日：2005-07-14
申请号：US10755374
申请日：2004-01-13
Applicant: Peter Stubley
Inventor： Peter Stubley
CPC classification number: H04M3/527 , G10L15/26 , H04M3/42059 , H04M3/42204 , H04M3/4931 , H04M2201/12 , H04M2201/40
Abstract: A method and apparatus for identifying a called party suitable for use in an automated attendant system are provided. Information derived from a spoken utterance by a caller is received. Identification information associated to the caller is derived. The information derived from the spoken utterance is processed on the basis of a plurality of directory entries to identify at least one directory entry that is a potential match to the information derived from the spoken utterance. When multiple directory entries in the plurality of directory entries are potential matches to the information, a calling pattern associated to the identification information is identified and a most likely directory entry from the multiple directory entries is selected at least in part on the basis of the calling pattern. A signal conveying the selected directory entry is then released.
Abstract translation: 提供了一种用于识别适合于在自动话务员系统中使用的被叫方的方法和设备。 收到来自呼叫者的口语发音的信息。 导出与呼叫者相关联的识别信息。 从多个目录条目的基础上处理从口语发音得到的信息，以识别至少一个目录条目，该目录条目是与从口头发音中导出的信息的潜在匹配。 当多个目录条目中的多个目录条目与信息是潜在的匹配时，识别与识别信息相关联的呼叫模式，并且至少部分地基于呼叫来选择来自多个目录条目的最可能的目录条目 模式。 然后释放传送所选目录条目的信号。
-
3.发明授权Method and apparatus for obtaining transcriptions from multiple training utterances 失效用于从多个训练语句获得转录的方法和装置
公开(公告)号：US5983177A
公开(公告)日：1999-11-09
申请号：US994007
申请日：1997-12-18
Applicant: Jianxiong Wu , Peter Stubley , Vishwa Gupta , Jean-Guy Dahan
Inventor： Jianxiong Wu , Peter Stubley , Vishwa Gupta , Jean-Guy Dahan
IPC: G10L5/06
CPC classification number: G10L15/06 , G10L15/187
Abstract: The invention relates to a method and an apparatus for adding a new entry to a speech recognition dictionary, more particularly to a system and method for generating transcriptions from multiple utterances of a given word. The novel method and apparatus automatically transcribes several training utterances into transcriptions without knowledge of the orthography of the word being added. It also provides a method and apparatus for transcribing multiple utterances into a single transcription that can be added to a speech recognition dictionary. In a first step, each utterance is analyzed individually to get their respective acoustic characteristics. Following this, these characteristics are combined to generate a set of the most likely transcriptions using the acoustic information obtained from each of the training utterances.
Abstract translation: 本发明涉及一种用于将新条目添加到语音识别词典的方法和装置，更具体地说，涉及一种用于从给定单词的多个话语生成转录的系统和方法。 该新颖的方法和设备在不知道所添加的单词的正字法的情况下，自动将多个训练语言转录成转录。 它还提供了一种用于将多个话语转录成可以添加到语音识别词典的单个转录中的方法和装置。 在第一步中，单独分析每个话语以获得它们各自的声学特性。 接下来，这些特征被组合以使用从每个训练话语获得的声学信息来产生一组最可能的转录。
-
公开(公告)号：US08239203B2
公开(公告)日：2012-08-07
申请号：US12423527
申请日：2009-04-14
Applicant: Peter Stubley , Oded Monzon , Alex Gorodetski , Douglas Peters
Inventor： Peter Stubley , Oded Monzon , Alex Gorodetski , Douglas Peters
CPC classification number: G10L15/08 , G10L2015/0631
Abstract: Adjusting confidence score thresholds is described for a speech recognition engine. The speech recognition engine is implemented in multiple computer processes functioning in a computer processor, and is characterized by an associated receiver operating characteristic (ROC) curve. A results confirmation process interprets user confirmation of speech recognition results within a given confidence score threshold to create a confirmed portion of the ROC curve for the speech recognition engine. A curve extension process extends the confirmed portion of the ROC curve by extrapolation of unconfirmed speech recognition results beyond the confidence score threshold to generate an extended ROC curve. A threshold adjustment process adjusts the confidence score threshold based on the extended ROC curve to meet target operating constraints for operating the speech recognition engine to perform automatic speech recognition of user speech inputs.
Abstract translation: 为语音识别引擎描述了调整置信度分数阈值。 语音识别引擎在计算机处理器中运行的多个计算机进程中实现，并且其特征在于相关联的接收器操作特性（ROC）曲线。 结果确认过程在给定的置信分数阈值内解释用户对语音识别结果的确认，以创建用于语音识别引擎的ROC曲线的确认部分。 曲线扩展过程通过外推未确认的语音识别结果超出置信分数阈值来扩展ROC曲线的确认部分，以产生扩展的ROC曲线。 阈值调整过程基于扩展的ROC曲线来调整置信分数阈值，以满足用于操作语音识别引擎以执行用户语音输入的自动语音识别的目标操作约束。
-
-
-
Search Results
4
records
(took 0.077 seconds)
