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. 发明授权

US08914283B2 System and method for unsupervised and active learning for automatic speech recognition 有权
标题翻译：用于自动语音识别的无监督和主动学习的系统和方法
公开(公告)号：US08914283B2
公开(公告)日：2014-12-16
申请号：US13959351
申请日：2013-08-05
申请人： AT&T Intellectual Property II, L.P.
发明人： Dilek Zeynep Hakkani-Tur , Giuseppe Riccardi
IPC分类号： G10L15/26
CPC分类号： G10L15/063 , G10L15/065 , G10L15/18 , G10L15/26 , G10L15/265
摘要： A system and method is provided for combining active and unsupervised learning for automatic speech recognition. This process enables a reduction in the amount of human supervision required for training acoustic and language models and an increase in the performance given the transcribed and un-transcribed data.
摘要翻译：提供一种系统和方法，用于组合自动语音识别的主动和无监督学习。该过程能够减少训练声学和语言模型所需的人力监督的数量，以及给出转录和未转录数据的性能的增加。

2. 发明授权

US08909529B2 Method and system for automatically detecting morphemes in a task classification system using lattices 有权
标题翻译：在使用格子的任务分类系统中自动检测语素的方法和系统
公开(公告)号：US08909529B2
公开(公告)日：2014-12-09
申请号：US14081234
申请日：2013-11-15
申请人： AT&T Intellectual Property II, L.P.
发明人： Giuseppe Riccardi
IPC分类号： G10L15/04
CPC分类号： G06F17/28 , G10L15/02 , G10L15/07 , G10L15/1822 , G10L15/187 , G10L15/24 , G10L25/27 , G10L2015/025
摘要： The invention concerns a method and corresponding system for building a phonotactic mode for domain independent speech recognition. The method may include recognizing phones from a user's input communication using a current phonotactic model, detecting morphemes (acoustic and/or non-acoustic) from the recognized phones, and outputting the detected morphemes for processing. The method also updates the phonotactic model with the detected morphemes and stores the new model in a database for use by the system during the next user interaction. The method may also include making task-type classification decisions based on the detected morphemes from the user's input communication.
摘要翻译：本发明涉及用于构建域独立语音识别的语音模式的方法和对应系统。该方法可以包括使用当前语音模型从用户的输入通信识别电话，从识别的电话中检测语素（声学和/或非声学），并输出检测到的语素进行处理。该方法还使用检测到的语素来更新语音模型，并将新模型存储在数据库中，以便在下一次用户交互期间由系统使用。该方法还可以包括基于来自用户的输入通信的检测到的语素来进行任务类型分类决定。

3. 发明申请

US20140149121A1 Method of Handling Frequently Asked Questions in a Natural Language Dialog Service 审中-公开
标题翻译：在自然语言对话服务中处理常见问题的方法
公开(公告)号：US20140149121A1
公开(公告)日：2014-05-29
申请号：US14171129
申请日：2014-02-03
申请人： AT&T Intellectual Property II, L.P.
发明人： Giuseppe Di Fabbrizio , Dawn L. Dutton , Narendra K. Gupta , Barbara B. Hollister , Mazin G. Rahim , Giuseppe Riccardi , Robert Elias Schapire , Juergen Schroeter
IPC分类号： G10L21/00
CPC分类号： G10L21/00 , G06F17/27 , G10L13/02
摘要： A voice-enabled help desk service is disclosed. The service comprises an automatic speech recognition module for recognizing speech from a user, a spoken language understanding module for understanding the output from the automatic speech recognition module, a dialog management module for generating a response to speech from the user, a natural voices text-to-speech synthesis module for synthesizing speech to generate the response to the user, and a frequently asked questions module. The frequently asked questions module handles frequently asked questions from the user by changing voices and providing predetermined prompts to answer frequently asked questions.
摘要翻译：公开了支持语音的帮助台服务。该服务包括用于识别来自用户的语音的自动语音识别模块，用于理解来自自动语音识别模块的输出的口语语言理解模块，用于生成来自用户对语音的响应的对话管理模块，自然语音文本 - 语音合成模块，用于合成语音以产生对用户的响应，以及常见问题模块。常见问题模块通过改变语音来处理用户的常见问题，并提供预定的提示来回答常见问题。

4. 发明授权

US09330660B2 Grammar fragment acquisition using syntactic and semantic clustering 有权
标题翻译：使用语法和语义聚类的语法片段获取
公开(公告)号：US09330660B2
公开(公告)日：2016-05-03
申请号：US14196536
申请日：2014-03-04
申请人： AT&T Intellectual Property II, L.P.
发明人： Kazuhiro Arai , Allen L. Gorin , Giuseppe Riccardi , Jeremy H. Wright
IPC分类号： G10L15/00 , G10L15/06 , G06F17/27 , G10L15/193
CPC分类号： G10L15/063 , G06F17/274 , G10L15/193 , G10L2015/0631
摘要： A method and apparatus are provided for automatically acquiring grammar fragments for recognizing and understanding fluently spoken language. Grammar fragments representing a set of syntactically and semantically similar phrases may be generated using three probability distributions: of succeeding words, of preceding words, and of associated call-types. The similarity between phrases may be measured by applying Kullback-Leibler distance to these tree probability distributions. Phrases being close in all three distances may be clustered into a grammar fragment.
摘要翻译：提供了一种方法和装置，用于自动获取用于识别和理解流利的口语的语法片段。可以使用三个概率分布来生成代表一组语法和语义上类似的短语的语法片段：前一个单词的后续单词和相关联的呼叫类型。可以通过将Kullback-Leibler距离应用于这些树概率分布来测量短语之间的相似性。所有三个距离中的短语可能被聚集成语法片段。

5. 发明申请

US20140074476A1 Method and System for Building a Phonotactic Model for Domain Independent Speech Recognition 有权
标题翻译：构建语言独立语音识别语音模型的方法与系统
公开(公告)号：US20140074476A1
公开(公告)日：2014-03-13
申请号：US14081234
申请日：2013-11-15
申请人： AT&T Intellectual Property II, L.P.
发明人： Giuseppe Riccardi
IPC分类号： G10L15/02
CPC分类号： G06F17/28 , G10L15/02 , G10L15/07 , G10L15/1822 , G10L15/187 , G10L15/24 , G10L25/27 , G10L2015/025
摘要： The invention concerns a method and corresponding system for building a phonotactic mode for domain independent speech recognition. The method may include recognizing phones from a user's input communication using a current phonotactic model detecting morphemes (acoustic and/or non-acoustic) from the recognized phones, and outputting die detected morphemes for processing. The method also updates the phonotactic model with the detected morphemes and stores the new model in a database for use by the system daring the next user interaction. The method may also include making task-type classification decisions based on the detected morphemes from the user's input communication.
摘要翻译：本发明涉及用于构建域独立语音识别的语音模式的方法和相应系统。该方法可以包括使用从识别的电话检测语素（声学和/或非声学）的当前语音模型识别来自用户的输入通信的电话，并且输出用于处理的检测到的词素。该方法还使用检测到的语素来更新语音模型，并将新模型存储在数据库中，以供系统大胆使用下一个用户交互。该方法还可以包括基于来自用户的输入通信的检测到的语素来进行任务类型分类决定。

6. 发明授权

US08612212B2 Method and system for automatically detecting morphemes in a task classification system using lattices 有权
标题翻译：在使用格子的任务分类系统中自动检测语素的方法和系统
公开(公告)号：US08612212B2
公开(公告)日：2013-12-17
申请号：US13784086
申请日：2013-03-04
申请人： AT&T Intellectual Property II, L.P.
发明人： Giuseppe Riccardi
IPC分类号： G06F17/27
CPC分类号： G06F17/28 , G10L15/02 , G10L15/07 , G10L15/1822 , G10L15/187 , G10L15/24 , G10L25/27 , G10L2015/025
摘要： The invention concerns a method and corresponding system for building a phonotactic model for domain independent speech recognition. The method may include recognizing phones from a user's input communication using a current phonotactic model, detecting morphemes (acoustic and/or non-acoustic) from the recognized phones, and outputting the detected morphemes for processing. The method also updates the phonotactic model with the detected morphemes and stores the new model in a database for use by the system during the next user interaction. The method may also include making task-type classification decisions based on the detected morphemes from the user's input communication.
摘要翻译：本发明涉及用于构建用于域独立语音识别的语音模型的方法和相应系统。该方法可以包括使用当前语音模型从用户的输入通信识别电话，从识别的电话中检测语素（声学和/或非声学），并输出检测到的语素进行处理。该方法还使用检测到的语素来更新语音模型，并将新模型存储在数据库中，以便在下一次用户交互期间由系统使用。该方法还可以包括基于来自用户的输入通信的检测到的语素来进行任务类型分类决定。

7. 发明申请

US20160027434A1 UNSUPERVISED AND ACTIVE LEARNING IN AUTOMATIC SPEECH RECOGNITION FOR CALL CLASSIFICATION 有权
公开(公告)号：US20160027434A1
公开(公告)日：2016-01-28
申请号：US14874843
申请日：2015-10-05
申请人： AT&T Intellectual Property II, L.P.
发明人： Dilek Z. Hakkani-Tur , Mazin G. Rahim , Giuseppe Riccardi , Gokhan Tur
IPC分类号： G10L15/06 , G10L15/26 , G10L15/18 , G10L15/07
CPC分类号： G10L15/063 , G10L15/07 , G10L15/18 , G10L15/26 , G10L2015/0638
摘要： Utterance data that includes at least a small amount of manually transcribed data is provided. Automatic speech recognition is performed on ones of the utterance data not having a corresponding manual transcription to produce automatically transcribed utterances. A model is trained using all of the manually transcribed data and the automatically transcribed utterances. A predetermined number of utterances not having a corresponding manual transcription are intelligently selected and manually transcribed. Ones of the automatically transcribed data as well as ones having a corresponding manual transcription are labeled. In another aspect of the invention, audio data is mined from at least one source, and a language model is trained for call classification from the mined audio data to produce a language model.

8. 发明申请

US20150046159A1 UNSUPERVISED AND ACTIVE LEARNING IN AUTOMATIC SPEECH RECOGNITION FOR CALL CLASSIFICATION 有权
标题翻译：自动语音识别中的不一致和主动学习用于呼叫分类
公开(公告)号：US20150046159A1
公开(公告)日：2015-02-12
申请号：US14468375
申请日：2014-08-26
申请人： AT&T Intellectual Property II, L.P.
发明人： Dilek Z. Hakkani-Tur , Mazin G. Rahim , Giuseppe Riccardi , Gokhan Tur
IPC分类号： G10L15/18 , G10L15/26
CPC分类号： G10L15/063 , G10L15/07 , G10L15/18 , G10L15/26 , G10L2015/0638
摘要： Utterance data that includes at least a small amount of manually transcribed data is provided. Automatic speech recognition is performed on ones of the utterance data not having a corresponding manual transcription to produce automatically transcribed utterances. A model is trained using all of the manually transcribed data and the automatically transcribed utterances. A predetermined number of utterances not having a corresponding manual transcription are intelligently selected and manually transcribed. Ones of the automatically transcribed data as well as ones having a corresponding manual transcription are labeled. In another aspect of the invention, audio data is mined from at least one source, and a language model is trained for call classification from the mined audio data to produce a language model.
摘要翻译：提供了至少包含少量手动转录数据的语音数据。对没有相应的手动转录的话语数据中的一个进行自动语音识别以产生自动转录的话语。使用所有手动转录数据和自动转录的话语训练模型。智能地选择并且手动地转录预定数量的不具有对应的手动转录的话语。自动转录的数据以及具有相应手动转录的数据的标签。在本发明的另一方面，音频数据从至少一个源开始，并且语言模型被训练用于从所开采的音频数据进行呼叫分类以产生语言模型。

9. 发明申请

US20140163988A1 Recognizing the Numeric Language in Natural Spoken Dialogue 有权
标题翻译：认识自然语言对话中的数字语言
公开(公告)号：US20140163988A1
公开(公告)日：2014-06-12
申请号：US14182017
申请日：2014-02-17
申请人： AT&T Intellectual Property II, L.P.
发明人： Mazin G. Rahim , Giuseppe Riccardi , Jeremy Huntley Wright , Bruce Melvin Buntschuh , Allen Louis Gorin
IPC分类号： G10L15/14
CPC分类号： G10L15/142
摘要： A system and a method are provided. A speech recognition processor receives unconstrained input speech and outputs a string of words. The speech recognition processor is based on a numeric language that represents a subset of a vocabulary. The subset includes a set of words identified as being for interpreting and understanding number strings. A numeric understanding processor contains classes of rules for converting the string of words into a sequence of digits. The speech recognition processor utilizes an acoustic model database. A validation database stores a set of valid sequences of digits. A string validation processor outputs validity information based on a comparison of a sequence of digits output by the numeric understanding processor with valid sequences of digits in the validation database.
摘要翻译：提供了一种系统和方法。语音识别处理器接收无约束输入语音并输出一串字。语音识别处理器基于代表词汇子集的数字语言。该子集包括被识别为用于解释和理解数字串的一组单词。数字理解处理器包含用于将字符串转换为数字序列的规则类型。语音识别处理器使用声学模型数据库。验证数据库存储一组有效的数字序列。字符串验证处理器基于数字理解处理器输出的数字序列与验证数据库中的有效数字序列的比较来输出有效性信息。

10. 发明申请

US20140156275A1 Method of Active Learning for Automatic Speech Recognition 有权
标题翻译：自动语音识别主动学习方法
公开(公告)号：US20140156275A1
公开(公告)日：2014-06-05
申请号：US14176439
申请日：2014-02-10
申请人： AT&T Intellectual Property II, L.P.
发明人： Allen Louis Gorin , Dilek Z. Hakkani-Tur , Giuseppe Riccardi
IPC分类号： G10L15/06
CPC分类号： G10L15/063
摘要： State-of-the-art speech recognition systems are trained using transcribed utterances, preparation of which is labor-intensive and time-consuming. The present invention is an iterative method for reducing the transcription effort for training in automatic speech recognition (ASR). Active learning aims at reducing the number of training examples to be labeled by automatically processing the unlabeled examples and then selecting the most informative ones with respect to a given cost function for a human to label. The method comprises automatically estimating a confidence score for each word of the utterance and exploiting the lattice output of a speech recognizer, which was trained on a small set of transcribed data. An utterance confidence score is computed based on these word confidence scores; then the utterances are selectively sampled to be transcribed using the utterance confidence scores.
摘要翻译：最先进的语音识别系统是使用转录语言进行训练，其准备是劳动密集型和耗时的。本发明是用于减少自动语音识别（ASR）中训练的转录努力的迭代方法。主动学习旨在通过自动处理未标记的示例，然后针对人类给定的成本函数选择最具信息的示例来减少要标注的训练示例的数量。该方法包括自动估计每个词语的置信度，并利用在一小组转录数据上训练的语音识别器的格子输出。基于这些单词置信度得出一个话语置信度得分; 然后使用话语置信度得分选择性地采样语音以进行转录。

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