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

US09406299B2 Differential acoustic model representation and linear transform-based adaptation for efficient user profile update techniques in automatic speech recognition 有权
标题翻译：差分声学模型表示和基于线性变换的自适应，用于自动语音识别中的高效用户简档更新技术
公开(公告)号：US09406299B2
公开(公告)日：2016-08-02
申请号：US14399867
申请日：2012-03-08
申请人： Christian Gollan , Daniel Willett
发明人： Christian Gollan , Daniel Willett
IPC分类号： G10L15/00 , G10L17/06 , G10L15/28 , G10L15/07 , G10L15/22 , G10L15/30
CPC分类号： G10L17/06 , G10L15/07 , G10L15/285 , G10L15/30 , G10L2015/227
摘要： A computer-implemented method is described for speaker adaptation in automatic speech recognition. Speech recognition data from a particular speaker is used for adaptation of an initial speech recognition acoustic model to produce a speaker adapted acoustic model. A speaker dependent differential acoustic model is determined that represents differences between the initial speech recognition acoustic model and the speaker adapted acoustic model. In addition, an approach is also disclosed to estimate speaker-specific feature or model transforms over multiple sessions. This is achieved by updating the previously estimated transform using only adaptation statistics of the current session.
摘要翻译：描述了用于自动语音识别中的说话者适应的计算机实现的方法。来自特定扬声器的语音识别数据用于初始语音识别声学模型的适应以产生扬声器适应的声学模型。确定说明者依赖的差分声学模型，其表示初始语音识别声学模型和说话者适应的声学模型之间的差异。此外，还公开了一种方法来估计多个会话中的说话人特定特征或模型变换。这是通过仅使用当前会话的适配统计来更新先前估计的变换来实现的。

2. 发明授权

US09037463B2 Efficient exploitation of model complementariness by low confidence re-scoring in automatic speech recognition 有权
标题翻译：通过自动语音识别中的低置信度重新评估模型互补性的高效利用
公开(公告)号：US09037463B2
公开(公告)日：2015-05-19
申请号：US13518409
申请日：2010-05-27
申请人： Daniel Willett , Venkatesh Nagesha
发明人： Daniel Willett , Venkatesh Nagesha
IPC分类号： G10L15/04 , G10L15/14 , G10L15/183 , G10L15/32
CPC分类号： G10L15/142 , G10L15/183 , G10L15/32
摘要： A method for speech recognition is described that uses an initial recognizer to perform an initial speech recognition pass on an input speech utterance to determine an initial recognition result corresponding to the input speech utterance, and a reliability measure reflecting a per word reliability of the initial recognition result. For portions of the initial recognition result where the reliability of the result is low, a re-evaluation recognizer is used to perform a re-evaluation recognition pass on the corresponding portions of the input speech utterance to determine a re-evaluation recognition result corresponding to the re-evaluated portions of the input speech utterance. The initial recognizer and the re-evaluation recognizer are complementary so as to make different recognition errors. A final recognition result is determined based on the re-evaluation recognition result if any, and otherwise based on the initial recognition result.
摘要翻译：描述了一种用于语音识别的方法，其使用初始识别器对输入语音话语执行初始语音识别，以确定对应于输入语音话语的初始识别结果，以及反映初始识别的每字可靠性的可靠性度量结果。对于结果可靠性低的初始识别结果的部分，使用重新评估识别器对输入语音话语的相应部分执行重新评估识别，以确定对应于重新评估输入语音话语的部分。初始识别器和重新评估识别器是互补的，以便产生不同的识别错误。根据重新评估识别结果（如果有的话），否则根据初始识别结果确定最终识别结果。

3. 发明授权

US08996368B2 Online maximum-likelihood mean and variance normalization for speech recognition 有权
标题翻译：语音识别的在线最大似然均值和方差归一化
公开(公告)号：US08996368B2
公开(公告)日：2015-03-31
申请号：US13518405
申请日：2010-02-22
申请人： Daniel Willett
发明人： Daniel Willett
IPC分类号： G10L15/00 , G10L15/02 , G10L15/20 , G10L15/08 , G10L15/34
CPC分类号： G10L19/0212 , G10L15/02 , G10L15/08 , G10L15/20 , G10L15/34 , G10L19/00
摘要： A feature transform for speech recognition is described. An input speech utterance is processed to produce a sequence of representative speech vectors. A time-synchronous speech recognition pass is performed using a decoding search to determine a recognition output corresponding to the speech input. The decoding search includes, for each speech vector after some first threshold number of speech vectors, estimating a feature transform based on the preceding speech vectors in the utterance and partial decoding results of the decoding search. The current speech vector is then adjusted based on the current feature transform, and the adjusted speech vector is used in a current frame of the decoding search.
摘要翻译：描述用于语音识别的特征变换。处理输入语音话语以产生代表性语音向量的序列。使用解码搜索来执行时间同步语音识别遍，以确定对应于语音输入的识别输出。解码搜索包括对于在一些第一阈值数量的语音矢量之后的每个语音矢量，基于解码搜索的发音和部分解码结果中的先前语音向量来估计特征变换。然后基于当前特征变换来调整当前语音向量，并且在解码搜索的当前帧中使用经调整的语音向量。

4. 发明申请

US20090138265A1 Joint Discriminative Training of Multiple Speech Recognizers 有权
标题翻译：多语音识别器联合辨别训练
公开(公告)号：US20090138265A1
公开(公告)日：2009-05-28
申请号：US11945048
申请日：2007-11-26
申请人： Daniel Willett , Chuang He
发明人： Daniel Willett , Chuang He
IPC分类号： G10L15/00
CPC分类号： G10L15/30 , G10L15/063
摘要： Adjusting model parameters is described for a speech recognition system that combines recognition outputs from multiple speech recognition processes. Discriminative adjustments are made to model parameters of at least one acoustic model based on a joint discriminative criterion over multiple complementary acoustic models to lower recognition word error rate in the system.
摘要翻译：对于组合来自多个语音识别过程的识别输出的语音识别系统，描述了调整模型参数。基于在多个互补声学模型上的联合判别标准来对至少一个声学模型的模型参数进行判别性调整，以降低系统中的识别字错误率。

5. 发明授权

US09953638B2 Meta-data inputs to front end processing for automatic speech recognition 有权
公开(公告)号：US09953638B2
公开(公告)日：2018-04-24
申请号：US14411236
申请日：2012-06-28
申请人： Daniel Willett , Karl Jonas Lööf , Yue Pan , Joel Pinto , Christian Gollan
发明人： Daniel Willett , Karl Jonas Lööf , Yue Pan , Joel Pinto , Christian Gollan
IPC分类号： G10L15/14 , G10L15/20 , G10L25/30 , G10L15/02
CPC分类号： G10L15/14 , G10L15/02 , G10L15/20 , G10L25/30
摘要： A computer-implemented method is described for front end speech processing for automatic speech recognition. A sequence of speech features which characterize an unknown speech input provided on an audio input channel and associated meta-data which characterize the audio input channel are received. The speech features are transformed with a computer process that uses a trained mapping function controlled by the meta-data, and automatic speech recognition is performed of the transformed speech features.

6. 发明申请

US20160027435A1 METHOD FOR TRAINING AN AUTOMATIC SPEECH RECOGNITION SYSTEM 审中-公开
标题翻译：用于训练自动语音识别系统的方法
公开(公告)号：US20160027435A1
公开(公告)日：2016-01-28
申请号：US14773142
申请日：2013-03-07
申请人： Joel PINTO , Josef Damianus ANASTASIADIS , Daniel WILLETT , NUANCE COMMUNICATIONS, INC.
发明人： Joel Pinto , Josef Damianus Anastasiadis , Daniel Willett
IPC分类号： G10L15/07 , G10L15/26 , G10L15/06
CPC分类号： G10L15/07 , G10L15/063 , G10L15/20 , G10L15/26 , G10L2015/0635 , G10L2015/0638
摘要： A system and method for speech recognition is provided. Embodiments may include receiving, at a first computing device, a far-talk signal from a far-talk computing device, the far-talk signal transmitted using a first channel and corresponding to an audible sound. Embodiments may further include receiving, at the first computing device, a near-talk signal from a near-talk computing device, the near-talk signal transmitted using a second channel and corresponding to the audible sound, wherein the far-talk signal and the near-talk signal are received during an enrollment phase of a far-talk speech recognition system. Embodiments may also include updating, at the first computing device, one or more models associated with a far-talk speech recognition system based upon, at least in part, one or more characteristics of the far-talk signal and one or more characteristics of the near-talk signal.
摘要翻译：提供了一种用于语音识别的系统和方法。实施例可以包括在第一计算设备处接收来自远程计算设备的远程通话信号，使用第一通道发送并对应于可听见的声音的远程通话信号。实施例还可以包括在第一计算设备处接收来自近距离通话计算设备的近距离通话信号，使用第二通道发送并对应于可听见的声音的近距离通话信号，其中远程通话信号和在远程通话语音识别系统的注册阶段期间接收近距离通话信号。实施例还可以包括基于至少部分地基于远程通话信号的一个或多个特性以及所述远程通话语音识别系统的一个或多个特征来更新与第一计算设备相关联的一个或多个模型近距离信号。

7. 发明申请

US20150262575A1 META-DATA INPUTS TO FRONT END PROCESSING FOR AUTOMATIC SPEECH RECOGNITION 有权
标题翻译：元数据输入到自动语音识别的前端处理
公开(公告)号：US20150262575A1
公开(公告)日：2015-09-17
申请号：US14411236
申请日：2012-06-28
申请人： Daniel Willett , Karl Jonas Lööf , Yue Pan , Joel Pinto , Christian Gollan
发明人： Daniel Willett , Karl Jonas Lööf , Yue Pan , Joel Pinto , Christian Gollan
IPC分类号： G10L15/14
CPC分类号： G10L15/14 , G10L15/02 , G10L15/20 , G10L25/30
摘要： A computer-implemented method is described for front end speech processing for automatic speech recognition. A sequence of speech features which characterize an unknown speech input provided on an audio input channel and associated meta-data data which characterize the audio input channel are received. The speech features are transformed with a computer process that uses a trained mapping function controlled by the meta-data, and automatic speech recognition is performed of the transformed speech features.
摘要翻译：描述了用于自动语音识别的前端语音处理的计算机实现的方法。接收表征提供在音频输入通道上的未知语音输入和表征音频输入通道的相关联的元数据数据的语音特征序列。通过使用由元数据控制的训练映射函数的计算机处理来转换语音特征，并且对转换的语音特征进行自动语音识别。

8. 发明授权

US08843370B2 Joint discriminative training of multiple speech recognizers 有权
标题翻译：多语音识别器的联合辨别训练
公开(公告)号：US08843370B2
公开(公告)日：2014-09-23
申请号：US11945048
申请日：2007-11-26
申请人： Daniel Willett , Chuang He
发明人： Daniel Willett , Chuang He
IPC分类号： G10L15/06 , G10L15/14 , G10L15/30
CPC分类号： G10L15/30 , G10L15/063
摘要： Adjusting model parameters is described for a speech recognition system that combines recognition outputs from multiple speech recognition processes. Discriminative adjustments are made to model parameters of at least one acoustic model based on a joint discriminative criterion over multiple complementary acoustic models to lower recognition word error rate in the system.
摘要翻译：对于组合来自多个语音识别过程的识别输出的语音识别系统，描述了调整模型参数。基于在多个互补声学模型上的联合判别标准来对至少一个声学模型的模型参数进行判别性调整，以降低系统中的识别字错误率。

9. 发明授权

US09984676B2 Feature normalization inputs to front end processing for automatic speech recognition 有权
公开(公告)号：US09984676B2
公开(公告)日：2018-05-29
申请号：US14414793
申请日：2012-07-24
申请人： Dermot Connolly , Daniel Willett
发明人： Dermot Connolly , Daniel Willett
IPC分类号： G10L15/02 , G10L15/24 , G10L25/24
CPC分类号： G10L15/02 , G10L15/24 , G10L25/24
摘要： A computer-implemented method is described for front end speech processing for automatic speech recognition. A sequence of speech features which characterize an unknown speech input is received with a computer process. A first subset of the speech features is normalized with a computer process using a first feature normalizing function. A second subset of the speech features is normalized with a computer process using a second feature normalizing function different from the first feature normalizing function. The normalized speech features in the first and second subsets are combined with a computer process to produce a sequence of mixed normalized speech features for automatic speech recognition.

10. 发明申请

US20150206527A1 FEATURE NORMALIZATION INPUTS TO FRONT END PROCESSING FOR AUTOMATIC SPEECH RECOGNITION 有权
标题翻译：特征正则化输入到自动语音识别的前端处理
公开(公告)号：US20150206527A1
公开(公告)日：2015-07-23
申请号：US14414793
申请日：2012-07-24
申请人： Dermot Connolly , Daniel Willett
发明人： Dermot Connolly , Daniel Willett
IPC分类号： G10L15/02 , G10L15/24
CPC分类号： G10L15/02 , G10L15/24 , G10L25/24
摘要： A computer-implemented method is described for front end speech processing for automatic speech recognition. A sequence of speech features which characterize an unknown speech input is received with a computer process. A first subset of the speech features is normalized with a computer process using a first feature normalizing function. A second subset of the speech features is normalized with a computer process using a second feature normalizing function different from the first feature normalizing function. The normalized speech features in the first and second subsets are combined with a computer process to produce a sequence of mixed normalized speech features for automatic speech recognition.
摘要翻译：描述了用于自动语音识别的前端语音处理的计算机实现的方法。用计算机进程接收表征未知语音输入的语音特征序列。通过使用第一特征归一化函数的计算机进程来对语音特征的第一子集进行归一化。使用不同于第一特征归一化函数的第二特征归一化函数的计算机处理对语音特征的第二子集进行归一化。第一和第二子集中的归一化语音特征与计算机过程组合以产生用于自动语音识别的混合规范化语音特征的序列。

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