公开(公告)号：US20190214020A1

公开(公告)日：2019-07-11

申请号：US16353756

申请日：2019-03-14

申请人： BEIJING DIDI INFINITY TECHNOLOGY AND DEVELOPMENT CO., LTD.

发明人： Jie CHEN ,  Dan SU ,  Tianxiao FU ,  Na HU

IPC分类号： G10L17/00 ,  G10L17/06 ,  G10L17/18 ,  G10L15/02 ,  G10L17/16

CPC分类号： G10L17/005 ,  G10L15/02 ,  G10L17/02 ,  G10L17/06 ,  G10L17/16 ,  G10L17/18 ,  G10L25/18 ,  G10L2025/783

摘要： The present disclosure relates to a method, apparatus, and system for speaker verification. The method includes: acquiring an audio recording; extracting speech signals from the audio recording; extracting features of the extracted speech signals; and determining whether the extracted speech signals represent speech by a predetermined speaker based on the extracted features and a speaker model trained with reference voice data of the predetermined speaker.

公开(公告)号：US09881617B2

公开(公告)日：2018-01-30

申请号：US15254326

申请日：2016-09-01

申请人： Verint Systems Ltd.

发明人： Oana Sidi ,  Ron Wein

IPC分类号： G10L15/26 ,  G10L17/06 ,  G10L17/02 ,  G10L17/16 ,  G10L15/02 ,  G10L17/04 ,  G10L25/78 ,  H04M3/51 ,  G10L15/00

CPC分类号： G10L17/06 ,  G10L15/02 ,  G10L15/26 ,  G10L17/02 ,  G10L17/04 ,  G10L17/16 ,  G10L25/78 ,  G10L2015/025 ,  H04M3/5175 ,  H04M2201/41 ,  H04M2203/303

摘要： In a method of diarization of audio data, audio data is segmented into a plurality of utterances. Each utterance is represented as an utterance model representative of a plurality of feature vectors. The utterance models are clustered. A plurality of speaker models are constructed from the clustered utterance models. A hidden Markov model is constructed of the plurality of speaker models. A sequence of identified speaker models is decoded.

公开(公告)号：US09875742B2

公开(公告)日：2018-01-23

申请号：US15006572

申请日：2016-01-26

申请人： Verint Systems Ltd.

发明人： Alex Gorodetski ,  Oana Sidi ,  Ron Wein ,  Ido Shapira

IPC分类号： G10L15/00 ,  G10L17/00 ,  G10L15/06 ,  G10L17/04 ,  G10L17/16 ,  G10L17/02 ,  G10L25/84 ,  G10L15/26

CPC分类号： G10L17/04 ,  G10L15/26 ,  G10L17/02 ,  G10L17/16 ,  G10L25/84

摘要： Disclosed herein are methods of diarizing audio data using first-pass blind diarization and second-pass blind diarization that generate speaker statistical models, wherein the first pass-blind diarization is on a per-frame basis and the second pass-blind diarization is on a per-word basis, and methods of creating acoustic signatures for a common speaker based only on the statistical models of the speakers in each audio session.

公开(公告)号：US20170053653A1

公开(公告)日：2017-02-23

申请号：US15254326

申请日：2016-09-01

申请人： Verint Systems Ltd.

发明人： Oana Sidi ,  Ron Wein

IPC分类号： G10L17/06 ,  G10L17/02 ,  H04M3/51 ,  G10L15/02 ,  G10L17/16 ,  G10L17/04 ,  G10L15/26 ,  G10L25/78

CPC分类号： G10L17/06 ,  G10L15/02 ,  G10L15/26 ,  G10L17/02 ,  G10L17/04 ,  G10L17/16 ,  G10L25/78 ,  G10L2015/025 ,  H04M3/5175 ,  H04M2201/41 ,  H04M2203/303

摘要： In a method of diarization of audio data, audio data is segmented into a plurality of utterances. Each utterance is represented as an utterance model representative of a plurality of feature vectors. The utterance models are clustered. A plurality of speaker models are constructed from the clustered utterance models. A hidden Markov model is constructed of the plurality of speaker models. A sequence of identified speaker models is decoded.

摘要翻译： 在音频数据的分类方法中，将音频数据分割为多个话语。 每个话语被表示为代表多个特征向量的话语模型。 话语模型是聚类的。 从群集话语模型构建多个说话者模型。 由多个扬声器模型构成隐马尔可夫模型。 已识别的扬声器模型的序列被解码。

公开(公告)号：US09461958B1

公开(公告)日：2016-10-04

申请号：US15050204

申请日：2016-02-22

申请人： Greenfly, Inc.

发明人： Shawn David Green ,  Daniel Brian Kirschner

IPC分类号： H04L12/58 ,  H04L29/08 ,  H04N21/475 ,  G06Q50/10 ,  G06Q10/10 ,  G06Q50/00 ,  H04L29/06 ,  H04N21/4788

CPC分类号： H04L51/32 ,  G06F17/30017 ,  G06F17/30029 ,  G06F17/30035 ,  G06F17/3005 ,  G06F17/30867 ,  G06K9/00255 ,  G06K9/00288 ,  G06Q10/101 ,  G06Q50/01 ,  G06Q50/10 ,  G10L17/005 ,  G10L17/06 ,  G10L17/16 ,  G10L17/22 ,  H04L65/4084 ,  H04L67/06 ,  H04L67/22 ,  H04L67/306 ,  H04N21/4756 ,  H04N21/4758 ,  H04N21/4784 ,  H04N21/4788

摘要： A content distribution facilitation system is described comprising configured servers and a network interface configured to interface with a plurality of terminals in a client server relationship and optionally with a cloud-based storage system. A request from a first source for content comprising content criteria is received, the content criteria comprising content subject matter. At least a portion of the content request content criteria is transmitted to a selected content contributor. If recorded content is received from the first content contributor, the first source is provided with access to the received recorded content. The recorded content may be transmitted via one or more networks to one or more destination devices. Optionally, a voice analysis and/or facial recognition engine are utilized to determine if the recorded content is from the first content contributor.

摘要翻译： 描述内容分发促进系统，其包括配置的服务器和被配置为以客户端服务器关系与可选地与基于云的存储系统与多个终端进行接口的网络接口。 接收来自包含内容标准的内容的来自第一来源的请求，内容标准包括内容主题。 内容请求内容标准的至少一部分被传送到选定的内容提供者。 如果从第一内容提供者接收到记录的内容，则向第一个来源提供对所接收的记录内容的访问。 记录的内容可以经由一个或多个网络发送到一个或多个目的地设备。 可选地，使用语音分析和/或面部识别引擎来确定所记录的内容是否来自第一内容贡献者。

公开(公告)号：US20160210986A1

公开(公告)日：2016-07-21

申请号：US14997179

申请日：2016-01-15

申请人： Lena Foundation

发明人： Terrance D. Paul ,  Dongxin D. Xu ,  Sharmistha Sarkar Gray ,  Umit Yapanel ,  Jill S. Gilkerson ,  Jeffrey A. Richards

IPC分类号： G10L25/63 ,  G10L15/14 ,  G10L15/07 ,  G10L15/02

CPC分类号： G10L25/63 ,  A61B5/16 ,  A61B5/4803 ,  A61B5/7264 ,  A61B2503/06 ,  G10L15/00 ,  G10L15/02 ,  G10L15/063 ,  G10L15/075 ,  G10L15/14 ,  G10L17/16 ,  G10L17/26 ,  G10L25/66 ,  G10L2015/022 ,  G10L2015/0631

摘要： A method of determining an emotion of an utterance. The method can include receiving the utterance at a processor-based device comprising an audio engine. The method also can include extracting emotion-related acoustic features from the utterance. The method additionally can include comparing the emotion-related acoustic features to a plurality of emotion models that are representative of emotions. The method further can include selecting a model from the plurality of emotion models based on the comparing the emotion-related acoustic features to the plurality of emotion models. The method additionally can include outputting the emotion of the utterance, wherein the emotion corresponds to the selected model. Other embodiments are provided.

摘要翻译： 一种确定话语情感的方法。 该方法可以包括在包括音频引擎的基于处理器的设备处接收话音。 该方法还可以包括从话语中提取与情绪相关的声学特征。 该方法还可以包括将情感相关的声学特征与代表情绪的多个情感模型进行比较。 该方法还可以包括基于将情感相关的声学特征与多个情感模型进行比较，从多个情感模型中选择模型。 该方法还可以包括输出话语的情感，其中情绪对应于所选择的模型。 提供其他实施例。

公开(公告)号：US20160203832A1

公开(公告)日：2016-07-14

申请号：US14997172

申请日：2016-01-15

申请人： Lena Foundation

发明人： Terrance D. Paul ,  Dongxin D. Xu ,  Sharmistha Sarkar Gray ,  Umit Yapanel ,  Jill S. Gilkerson ,  Jeffrey A. Richards

IPC分类号： G10L25/66 ,  A61B5/16 ,  A61B5/00 ,  G10L15/02 ,  G10L15/06

CPC分类号： G10L25/63 ,  A61B5/16 ,  A61B5/4803 ,  A61B5/7264 ,  A61B2503/06 ,  G10L15/00 ,  G10L15/02 ,  G10L15/063 ,  G10L15/075 ,  G10L15/14 ,  G10L17/16 ,  G10L17/26 ,  G10L25/66 ,  G10L2015/022 ,  G10L2015/0631

摘要： In some embodiments, a method of creating an automatic language characteristic recognition system. The method can include receiving a plurality of audio recordings. The method also can include segmenting each of the plurality of audio recordings to create a plurality of audio segments for each audio recording. The method additionally can include clustering each audio segment of the plurality of audio segments according to audio characteristics of each audio segment to form a plurality of audio segment clusters. Other embodiments are provided.

摘要翻译： 在一些实施例中，创建自动语言特征识别系统的方法。 该方法可以包括接收多个音频记录。 该方法还可以包括分割多个音频记录中的每一个，以便为每个音频记录创建多个音频段。 该方法还可以包括根据每个音频片段的音频特性对多个音频片段中的每个音频片段进行聚类，以形成多个音频片段集群。 提供其他实施例。

公开(公告)号：US09240188B2

公开(公告)日：2016-01-19

申请号：US12359124

申请日：2009-01-23

申请人： Terrance D. Paul ,  Dongxin D. Xu ,  Sharmistha S. Gray ,  Umit Yapanel ,  Jill S. Gilkerson ,  Jeffrey A. Richards

发明人： Terrance D. Paul ,  Dongxin D. Xu ,  Sharmistha S. Gray ,  Umit Yapanel ,  Jill S. Gilkerson ,  Jeffrey A. Richards

IPC分类号： G10L17/26 ,  G10L15/02 ,  G10L15/00 ,  G10L17/16

CPC分类号： G10L25/63 ,  A61B5/16 ,  A61B5/4803 ,  A61B5/7264 ,  A61B2503/06 ,  G10L15/00 ,  G10L15/02 ,  G10L15/063 ,  G10L15/075 ,  G10L15/14 ,  G10L17/16 ,  G10L17/26 ,  G10L25/66 ,  G10L2015/022 ,  G10L2015/0631

摘要： In one embodiment, the system and method for expressive language development; a method for detecting autism in a natural language environment using a microphone, sound recorder, and a computer programmed with software for the specialized purpose of processing recordings captured by the microphone and sound recorder combination; and the computer programmed to execute a method that includes segmenting an audio signal captured by the microphone and sound recorder combination using the computer programmed for the specialized purpose into a plurality recording segments. The method further includes determining which of the plurality of recording segments correspond to a key child. The method also includes extracting acoustic parameters of the key child recordings and comparing the acoustic parameters of the key child recordings to known acoustic parameters for children. The method returns a determination of a likelihood of autism.

摘要翻译： 在一个实施例中，用于表达语言发展的系统和方法; 用于使用麦克风，录音机和用于处理由麦克风和录音机组合拍摄的记录的专门目的的软件编程的用于在自然语言环境中检测自闭症的方法; 并且所述计算机被编程为执行一种方法，该方法包括使用为专用目的而编程的计算机将由麦克风捕获的音频信号和声音记录器组合分割成多个记录段。 所述方法还包括确定所述多个记录段中的哪一个与密钥子对应。 该方法还包括提取关键子记录的声学参数并将关键子记录的声学参数与儿童的已知声学参数进行比较。 该方法返回自闭症的可能性的确定。

公开(公告)号：US20150025887A1

公开(公告)日：2015-01-22

申请号：US14319860

申请日：2014-06-30

申请人： VERINT SYSTEMS LTD.

发明人： Oana Sidi ,  Ron Wein

IPC分类号： G10L15/06

CPC分类号： G10L17/06 ,  G10L15/02 ,  G10L15/26 ,  G10L17/02 ,  G10L17/04 ,  G10L17/16 ,  G10L25/78 ,  G10L2015/025 ,  H04M3/5175 ,  H04M2201/41 ,  H04M2203/303

摘要： In a method of diarization of audio data, audio data is segmented into a plurality of utterances. Each utterance is represented as an utterance model representative of a plurality of feature vectors. The utterance models are clustered. A plurality of speaker models are constructed from the clustered utterance models. A hidden Markov model is constructed of the plurality of speaker models. A sequence of identified speaker models is decoded.

摘要翻译： 在音频数据的分类方法中，将音频数据分割为多个话语。 每个话语被表示为代表多个特征向量的话语模型。 话语模型是聚类的。 从群集话语模型构建多个说话者模型。 由多个扬声器模型构成隐马尔可夫模型。 已识别的扬声器模型的序列被解码。

公开(公告)号：US20140172428A1

公开(公告)日：2014-06-19

申请号：US14016611

申请日：2013-09-03

申请人： Electronics and Telecommunications Research Institute

发明人： Mun Sung HAN

IPC分类号： G10L17/16

CPC分类号： G10L17/16

摘要： Provided is a method for context independent gender recognition utilizing phoneme transition probability. The method for the context independent gender recognition includes detecting a voice section from a received voice signal, generating feature vectors within the detected voice section, performing a hidden Markov model on the feature vectors by using a search network that is set according to a phoneme rule to recognize a phoneme and obtain scores of first and second likelihoods, and comparing final scores of the first and second likelihoods obtained while the phoneme recognition is performed up to the last section of the voice section to finally decide gender with respect to the voice signal.

摘要翻译： 提供了利用音素转换概率的上下文独立性别识别的方法。 用于上下文独立性别识别的方法包括从接收到的语音信号中检测语音部分，在检测到的语音部分内生成特征向量，通过使用根据音素规则设置的搜索网络对特征向量执行隐马尔可夫模型 识别音素并获得第一和第二可能性的分数，并且将执行音素识别时获得的第一和第二可能性的最终分数与语音部分的最后部分进行比较，以最终确定关于语音信号的性别。