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

US20050027527A1 System and method enabling acoustic barge-in 有权
标题翻译：允许声学加载的系统和方法
公开(公告)号：US20050027527A1
公开(公告)日：2005-02-03
申请号：US10631985
申请日：2003-07-31
申请人： Jochen Junkawitsch , Raymond Bruckner , Klaus Reinhard , Stefan Dobler
发明人： Jochen Junkawitsch , Raymond Bruckner , Klaus Reinhard , Stefan Dobler
IPC分类号： G10L15/20 , G10L15/22 , G10L21/02 , G10L15/06
CPC分类号： G10L15/22 , G10L15/20 , G10L2021/02082
摘要： A system and method enabling acoustic barge-in during a voice prompt in a communication system. An acoustic prompt model is trained to represent the system prompt using the specific speech signal of the prompt. The acoustic prompt model is utilized in a speech recognizer in parallel with the recognizer's active vocabulary words to suppress the echo of the prompt within the recognizer. The speech recognizer may also use a silence model and traditional garbage models such as noise models and out-of-vocabulary word models to reduce the likelihood that noises and out-of-vocabulary words in the user utterance will be mapped erroneously onto active vocabulary words.
摘要翻译：一种在通信系统中的语音提示期间使声音插入的系统和方法。使用提示的特定语音信号训练声学提示模型来表示系统提示。声学提示模型与识别器的活动词汇单词并行地用于语音识别器，以抑制识别器内的提示的回波。语音识别器还可以使用静音模型和诸如噪声模型和词外词语模型之类的传统垃圾模型来减少用户话语中的噪声和词汇外词在错误地映射到活动词汇单词上的可能性。

2. 发明授权

US07392188B2 System and method enabling acoustic barge-in 有权
标题翻译：允许声学加载的系统和方法
公开(公告)号：US07392188B2
公开(公告)日：2008-06-24
申请号：US10631985
申请日：2003-07-31
申请人： Jochen Junkawitsch , Raymond Bruckner , Klaus Reinhard , Stefan Dobler
发明人： Jochen Junkawitsch , Raymond Bruckner , Klaus Reinhard , Stefan Dobler
IPC分类号： G10L15/06
CPC分类号： G10L15/22 , G10L15/20 , G10L2021/02082
摘要： A system and method enabling acoustic barge-in during a voice prompt in a communication system. An acoustic prompt model is trained to represent the system prompt using the specific speech signal of the prompt. The acoustic prompt model is utilized in a speech recognizer in parallel with the recognizer's active vocabulary words to suppress the echo of the prompt within the recognizer. The speech recognizer may also use a silence model and traditional garbage models such as noise models and out-of-vocabulary word models to reduce the likelihood that noises and out-of-vocabulary words in the user utterance will be mapped erroneously onto active vocabulary words.
摘要翻译：一种在通信系统中的语音提示期间使声音插入的系统和方法。使用提示的特定语音信号训练声学提示模型来表示系统提示。声学提示模型与识别器的活动词汇单词并行地用于语音识别器，以抑制识别器内的提示的回波。语音识别器还可以使用静音模型和诸如噪声模型和词外词语模型之类的传统垃圾模型来减少用户话语中的噪声和词汇外词在错误地映射到活动词汇单词上的可能性。

3. 发明申请

US20050038652A1 Method and device for voice recognition 有权
标题翻译：用于语音识别的方法和装置
公开(公告)号：US20050038652A1
公开(公告)日：2005-02-17
申请号：US10496769
申请日：2001-12-21
申请人： Stefan Dobler
发明人： Stefan Dobler
IPC分类号： G10L15/22 , G10L25/87 , G10L15/04
CPC分类号： G10L25/87 , G10L15/22
摘要： Method and device for the recognition of words and pauses in a voice signal. The words (Wi) spoken in a row and pauses (Ti) are thereby combined as to be appertaining to a word group as soon as one of the pauses (Ti) exceeds a limit value (TG). Stored references (Rj) are allocated to the voice signal of the word group, and an indication of the result of the allocation is effected after the limit value (TG) has been exceeded. To this end, parameters corresponding to the moments of the transitions between ranges with voice and non-voice are determined from the voice signal, and the limit value (TG) is then changed in dependence on said parameters.
摘要翻译：用于识别语音信号中的单词和暂停的方法和设备。因此，一旦暂停（Ti）超过限制值（TG），一行中的单词（Wi）和暂停（Ti）就被组合成与单词组相关。存储的参考（Rj）被分配给单词组的语音信号，并且在超过限制值（TG）之后实现分配结果的指示。为此，从语音信号确定与语音和非语音的范围之间的转换的时刻对应的参数，然后根据所述参数改变极限值（TG）。

4. 发明授权

US5784454A Arrangement for communication with a subscriber 失效
标题翻译：与订户进行通信的安排
公开(公告)号：US5784454A
公开(公告)日：1998-07-21
申请号：US508285
申请日：1995-07-27
申请人： Peter Patrick , Hans-Wilhelm Ruhl , Stefan Dobler
发明人： Peter Patrick , Hans-Wilhelm Ruhl , Stefan Dobler
IPC分类号： H04M1/60 , H04B3/21 , H04B3/23 , H04M9/00
CPC分类号： H04B3/21
摘要： The invention relates to an arrangement for communication with a subscriber. The arrangement includes a spectral analysis unit for producing short-time spectral values (Y(i)) of received signals (E) which signals are at times subscriber's speech signals superimposed by echoes of transmission signals (S) transmitted to the subscriber, a echo cancelling unit for estimating short-time spectral values of the echoes (X.sub.w (i)) and for producing difference values (D(i)) between the short-time spectral values (Y(i)) of the received signals (E) and the estimated short-time spectral values (X.sub.w (i)) of the echoes, and speech recognition unit for evaluating the difference values (D(i)).
摘要翻译：本发明涉及一种用于与用户通信的装置。该装置包括用于产生接收信号（E）的短时频谱值（Y（i））的频谱分析单元，该信号有时是由发送给用户的发送信号（S）的回波叠加的用户语音信号，回波消除单元，用于估计回波（Xw（i））的短时频谱值，并产生接收信号（E）和（E）的短时频谱值（Y（i））之间的差值回波的估计短时频谱值（Xw（i））和用于评估差值（D（i））的语音识别单元。

5. 发明授权

US06961702B2 Method and device for generating an adapted reference for automatic speech recognition 有权
标题翻译：用于生成用于自动语音识别的适配参考的方法和装置
公开(公告)号：US06961702B2
公开(公告)日：2005-11-01
申请号：US09985844
申请日：2001-11-06
申请人： Stefan Dobler , Andreas Kiessling , Ralph Schleifer , Raymond Brückner
发明人： Stefan Dobler , Andreas Kiessling , Ralph Schleifer , Raymond Brückner
IPC分类号： G10L15/06 , G10L15/07 , G10L15/22 , G10L15/10 , G10L15/12
CPC分类号： G10L15/07 , G10L2015/0631 , G10L2015/0635 , G10L2015/0636
摘要： The invention relates to a method for generating an adapted reference for automatic speech recognition. In a first step, recognition is performed based on a spoken utterance and a recognition result which corresponds to a currently valid reference is obtained. In a second step, the currently valid reference is adapted in accordance with the utterance in order to create an adapted reference. In a third step, the adapted reference is assessed and it is decided if the adapted reference is used for further recognition.
摘要翻译：本发明涉及一种用于生成用于自动语音识别的适配参考的方法。在第一步骤中，基于语音发音执行识别，并获得对应于当前有效参考的识别结果。在第二步中，为了创建适应的参考，当前有效的参考根据话语进行调整。在第三步中，评估适应参考，并且确定是否使用适配参考用于进一步识别。

6. 发明授权

US07366667B2 Method and device for pause limit values in speech recognition 有权
标题翻译：语音识别中暂停极限值的方法和装置
公开(公告)号：US07366667B2
公开(公告)日：2008-04-29
申请号：US10496769
申请日：2001-12-21
申请人： Stefan Dobler
发明人： Stefan Dobler
IPC分类号： G10L15/04 , G10L15/12 , G10L11/02
CPC分类号： G10L25/87 , G10L15/22
摘要： Method and device for the recognition of words and pauses in a voice signal. The words (Wi) spoken in a row and pauses (Ti) are thereby combined as to be appertaining to a word group as soon as one of the pauses (Ti) exceeds a limit value (TG). Stored references (Rj) are allocated to the voice signal of the word group, and an indication of the result of the allocation is effected after the limit value (TG) has been exceeded. To this end, parameters corresponding to the moments of the transitions between ranges with voice and non-voice are determined from the voice signal, and the limit value (TG) is then changed in dependence on said parameters.
摘要翻译：用于识别语音信号中的单词和暂停的方法和设备。因此，一旦暂停（T（T））中的一个暂停（T < 超过极限值（TG）。存储的引用（R SUB）被分配给字组的语音信号，并且在超过限制值（TG）之后进行分配结果的指示。为此，从语音信号确定与语音和非语音的范围之间的转换的时刻对应的参数，然后根据所述参数改变极限值（TG）。

7. 发明授权

US5687288A System with speaking-rate-adaptive transition values for determining words from a speech signal 失效
标题翻译：具有用于确定来自语音信号的单词的讲话速率自适应转换值的系统
公开(公告)号：US5687288A
公开(公告)日：1997-11-11
申请号：US528289
申请日：1995-09-14
申请人： Stefan Dobler , Hans-Wilhelm Ruhl
发明人： Stefan Dobler , Hans-Wilhelm Ruhl
IPC分类号： G10L15/12 , G10L15/06 , G10L5/06 , G10L9/06
CPC分类号： G10L15/07 , G10L15/12
摘要： Speech recognition produces test signals from the speech signal which are compared with predetermined reference signals so as to form scores. Each subsequent test signal is compared with reference values which are situated within a predetermined neighborhood of the reference value which has been determined to be optimum for the preceding test signal. In dependence on this neighborhood, transition values in conformity with the transition probabilities are added to the scores. In order to enhance the results notably in the case of different speeds of speaking of the instantaneous speaker, it is proposed to adapt these transition values in dependence on the speed of speaking. A further improvement can be achieved by also adapting the reference values themselves to the relevant speaker's pronunciation. This adaptation can also be iteratively performed in a number of steps.
摘要翻译：语音识别从语音信号产生与预定参考信号进行比较的测试信号，以形成分数。将每个后续测试信号与参考值进行比较，该参考值位于已经被确定为对于先前测试信号是最佳的参考值的预定邻域内。根据这个邻域，符合转移概率的转换值被添加到分数。为了在不同速度的瞬时扬声器的情况下显着提高结果，提出了根据演讲速度来适应这些转换值。还可以通过将参考值本身适应于相关说话者的发音来实现进一步的改进。这种适应也可以在多个步骤中迭代地执行。

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