公开(公告)号：US06513004B1

公开(公告)日：2003-01-28

申请号：US09449053

申请日：1999-11-24

申请人： Luca Rigazio ,  David Kryze ,  Ted Applebaum ,  Jean-Claude Junqua

发明人： Luca Rigazio ,  David Kryze ,  Ted Applebaum ,  Jean-Claude Junqua

IPC分类号： G10L1504

CPC分类号： G10L15/02 ,  G10L25/27

摘要： The acoustic speech signal is decomposed into wavelets arranged in an asymmetrical tree data structure from which individual nodes may be selected to best extract local features, as needed to model specific classes of sound units. The wavelet packet transformation is smoothed through integration and compressed to apply a non-linearity prior to discrete cosine transformation. The resulting subband features such as cepstral coefficients may then be used to construct the speech recognizer's speech models. Using the local feature information extracted in this manner allows a single recognizer to be optimized for several different classes of sound units, thereby eliminating the need for parallel path recognizers.

摘要翻译： 声学语音信号被分解成以不对称树数据结构排列的小波，根据需要可以选择各个节点以最佳地提取局部特征，以模拟特定类别的声音单元。 小波包变换通过积分进行平滑，并进行压缩，以在离散余弦变换之前应用非线性。 所得到的子带特征如倒谱系数可以用于构建语音识别器的语音模型。 使用以这种方式提取的局部特征信息允许为多个不同类别的声音单元优化单个识别器，从而消除对并行路径识别器的需要。

公开(公告)号：US07729908B2

公开(公告)日：2010-06-01

申请号：US11369936

申请日：2006-03-06

申请人： Luca Rigazio ,  David Kryze ,  Keiko Morii ,  Nobuyuki Kunieda ,  Jean-Claude Junqua

发明人： Luca Rigazio ,  David Kryze ,  Keiko Morii ,  Nobuyuki Kunieda ,  Jean-Claude Junqua

IPC分类号： G10L15/20 ,  G10L15/06 ,  G10L21/02

CPC分类号： G10L15/20 ,  G10L21/0216

摘要： A noise robustness method operates jointly in a signal domain and a model domain. For example, energy is added in the signal domain for frequency bands where an actual noise level of an incoming signal is lower than a noise level used to train models, thus obtaining a compensated signal. Also, energy is added in the model domain for frequency bands where noise level of the incoming signal or the compensated signal is higher than the noise level used to train the models. Moreover, energy is never removed, thereby avoiding problems of higher sensitivity of energy removal to estimation errors.

摘要翻译： 噪声鲁棒性方法在信号域和模型域中共同操作。 例如，在信号域中增加能量，其中输入信号的实际噪声电平低于用于训练模型的噪声电平，从而获得补偿信号。 此外，在模型域中增加能量，其中输入信号或补偿信号的噪声电平高于用于训练模型的噪声电平的频带。 此外，能量永远不会被去除，从而避免了能量去除对估计误差的更高灵敏度的问题。

公开(公告)号：US20110043475A1

公开(公告)日：2011-02-24

申请号：US12988745

申请日：2009-04-21

申请人： Luca Rigazio ,  David Kryze ,  Jean-Claude Junqua

发明人： Luca Rigazio ,  David Kryze ,  Jean-Claude Junqua

IPC分类号： G06F3/041

CPC分类号： H04N5/4403 ,  G06F3/04883 ,  G08C17/00 ,  G08C2201/30 ,  H04N21/42208 ,  H04N21/42222 ,  H04N21/42224 ,  H04N2005/443

摘要： A system and method for identifying a user of a handheld device is herein disclosed. The device implementing the method and system may attempt to identify a user based on signals that are incidental to a user's handling of the device. The signals are generated by a variety of sensors dispersed along the periphery or within the housing. The sensors range may include touch sensors, inertial sensors, acoustic sensors, pulse oximiters, and a touchpad. Based on the sensors and corresponding signals, identification information is generated. The identification information is used to identify the user of the handheld device. The handheld device may implement various statistical learning and data mining techniques to increase the robustness of the system. The device may also authenticate the user based on the user drawing a circle, or other shape.

摘要翻译： 本文公开了一种用于识别手持设备的用户的系统和方法。 实现该方法和系统的设备可以基于用户对设备的处理附带的信号来尝试识别用户。 信号由沿着周边或壳体内分散的各种传感器产生。 传感器范围可以包括触摸传感器，惯性传感器，声学传感器，脉冲嗅觉器和触摸板。 基于传感器和相应的信号，生成识别信息。 识别信息用于识别手持设备的用户。 手持设备可以实现各种统计学习和数据挖掘技术以增加系统的鲁棒性。 设备还可以基于用户绘制圆形或其他形状来认证用户。

公开(公告)号：US07729909B2

公开(公告)日：2010-06-01

申请号：US11369938

申请日：2006-03-06

申请人： Luca Rigazio ,  David Kryze ,  Keiko Morii ,  Nobuyuki Kunieda ,  Jean-Claude Junqua

发明人： Luca Rigazio ,  David Kryze ,  Keiko Morii ,  Nobuyuki Kunieda ,  Jean-Claude Junqua

IPC分类号： G10L15/20 ,  G10L15/06

CPC分类号： G10L15/20 ,  G10L15/065

摘要： Model compression is combined with model compensation. Model compression is needed in embedded ASR to reduce the size and the computational complexity of compressed models. Model-compensation is used to adapt in real-time to changing noise environments. The present invention allows for the design of smaller ASR engines (memory consumption reduced to up to one-sixth) with reduced impact on recognition accuracy and/or robustness to noises.

摘要翻译： 模型压缩与模型补偿相结合。 嵌入式ASR需要模型压缩，以减小压缩模型的大小和计算复杂度。 模型补偿用于实时改变噪声环境。 本发明允许将更小的ASR引擎（存储器消耗减少到高达六分之一）的设计减少，对识别精度和/或对噪声的鲁棒性的影响减小。

公开(公告)号：US06915259B2

公开(公告)日：2005-07-05

申请号：US09864838

申请日：2001-05-24

申请人： Luca Rigazio ,  Patrick Nguyen ,  David Kryze ,  Jean-Claude Junqua

发明人： Luca Rigazio ,  Patrick Nguyen ,  David Kryze ,  Jean-Claude Junqua

IPC分类号： G10L15/06 ,  G10L21/02

CPC分类号： G10L15/07 ,  G10L21/0208

摘要： Linear approximation of the background noise is applied after feature extraction and prior to speaker adaptation to allow the speaker adaptation system to adapt the speech models to the enrolling user without distortion from background noise. The linear approximation is applied in the feature domain, such as in the cepstral domain. Any adaptation technique that is commutative in the feature domain may be used.

摘要翻译： 背景噪声的线性近似在特征提取之后并且在说话者适配之前被应用，以允许扬声器适配系统将语音模型适应于登记用户，而不会从背景噪声失真。 线性近似应用于特征域，如倒谱域。 可以使用在特征域中可交换的任何适配技术。

公开(公告)号：US06907397B2

公开(公告)日：2005-06-14

申请号：US10245727

申请日：2002-09-16

申请人： David Kryze ,  Luca Rigazio ,  Patrick Nguyen ,  Jean-Claude Junqua

发明人： David Kryze ,  Luca Rigazio ,  Patrick Nguyen ,  Jean-Claude Junqua

IPC分类号： G10L15/00 ,  G06F17/30 ,  G10L20060101 ,  G10L11/00 ,  G10L15/04 ,  G10L15/06 ,  G10L15/18 ,  G10L15/26 ,  G10L21/00

CPC分类号： G06F17/30026 ,  G10L15/183 ,  G10L15/19 ,  G10L15/26 ,  Y10S707/99933 ,  Y10S707/99934 ,  Y10S707/99935

摘要： An embedded device for playing media files is capable of generating a play list of media files based on input speech from a user. It includes an indexer generating a plurality of speech recognition grammars. According to one aspect of the invention, the indexer generates speech recognition grammars based on contents of a media file header of the media file. According to another aspect of the invention, the indexer generates speech recognition grammars based on categories in a file path for retrieving the media file to a user location. When a speech recognizer receives an input speech from a user while in a selection mode, a media file selector compares the input speech received while in the selection mode to the plurality of speech recognition grammars, thereby selecting the media file.

摘要翻译： 用于播放媒体文件的嵌入式设备能够基于来自用户的输入语音来生成媒体文件的播放列表。 它包括产生多个语音识别语法的索引器。 根据本发明的一个方面，索引器基于媒体文件的媒体文件头的内容生成语音识别语法。 根据本发明的另一方面，索引器基于用于将媒体文件检索到用户位置的文件路径中的类别来生成语音识别语法。 当语音识别器在选择模式下从用户接收到输入语音时，媒体文件选择器将选择模式下接收到的输入语音与多个语音识别语法进行比较，从而选择媒体文件。

公开(公告)号：US20060009974A1

公开(公告)日：2006-01-12

申请号：US10888916

申请日：2004-07-09

申请人： Jean-Claude Junqua ,  Luca Rigazio ,  Jia Lei

发明人： Jean-Claude Junqua ,  Luca Rigazio ,  Jia Lei

IPC分类号： G10L15/18

CPC分类号： G10L15/083 ,  G10L2015/227 ,  G10L2015/228 ,  H04M1/271 ,  H04M2250/60

摘要： Dynamically constructed grammar-constraints and frequency or statistics-based constraints are used to constrain the speech recognizer and to optionally rescore the output to improve recognition accuracy. The recognition system is well adapted for hands-free operation of portable devices, such as for voice dialing operations.

摘要翻译： 动态构造的语法约束和基于频率或基于统计的约束用于限制语音识别器，并且可选地重新输出输出以提高识别精度。 识别系统很好地适用于便携式设备的免提操作，例如语音拨号操作。

公开(公告)号：US20050075881A1

公开(公告)日：2005-04-07

申请号：US10677174

申请日：2003-10-02

申请人： Luca Rigazio ,  Robert Boman ,  Patrick Nguyen ,  Jean-Claude Junqua

发明人： Luca Rigazio ,  Robert Boman ,  Patrick Nguyen ,  Jean-Claude Junqua

IPC分类号： G10L15/26 ,  G10L21/00

CPC分类号： G06F17/30796 ,  G10L15/26

摘要： A media capture device has an audio input receptive of user speech relating to a media capture activity in close temporal relation to the media capture activity. A plurality of focused speech recognition lexica respectively relating to media capture activities are stored on the device, and a speech recognizer recognizes the user speech based on a selected one of the focused speech recognition lexica. A media tagger tags captured media with generated speech recognition text, and a media annotator annotates the captured media with a sample of the user speech that is suitable for input to a speech recognizer. Tagging and annotating are based on close temporal relation between receipt of the user speech and capture of the captured media. Annotations may be converted to tags during post processing, employed to edit a lexicon using letter-to-sound rules and spelled word input, or matched directly to speech to retrieve captured media.

摘要翻译： 媒体捕获设备具有接收与媒体捕获活动紧密相关的媒体捕获活动的用户语音的音频输入。 分别与媒体捕获活动相关的多个聚焦语音识别词典被存储在设备上，并且语音识别器基于所选择的一个焦点语音识别词典识别用户语音。 媒体标签器使用生成的语音识别文本来标记捕获的媒体，并且媒体注释器用适合于输入到语音识别器的用户语音的样本来注释所捕获的媒体。 标记和注释是基于用户语音的接收和捕获的媒体的捕获之间的紧密的时间关系。 在后期处理中，注释可以转换为标签，用于使用字母对声音规则和拼写单词输入来编辑词典，或直接与语音匹配以检索所捕获的媒体。

公开(公告)号：US06901364B2

公开(公告)日：2005-05-31

申请号：US09951093

申请日：2001-09-13

申请人： Patrick Nguyen ,  Luca Rigazio ,  Jean-Claude Junqua

发明人： Patrick Nguyen ,  Luca Rigazio ,  Jean-Claude Junqua

IPC分类号： G10L15/18 ,  G10L15/28 ,  G10L15/26 ,  G06F17/20 ,  G10L21/00

CPC分类号： G10L15/1815 ,  G10L15/30

摘要： An e-mail message process is provided for use with a personal digital assistant which allows for the use of input speech messaging which is converted to text using a focused language model which is downloaded by a cellular phone connection to an Internet server which provides the focused language model based upon a topic for the intended e-mail message. The text that is generated from the input speech method can be summarized by the e-mail message processor and can be edited by the user. The generated e-mail message can then be transmitted again via cellular connection to an Internet e-mail server for transmitting the e-mail message to a recipient.

摘要翻译： 提供电子邮件消息处理以与个人数字助理一起使用，该个人数字助理允许使用输入语音消息传送，其使用由通过蜂窝电话连接下载的聚焦语言模型转换为文本，该互联网服务器提供聚焦 基于预期电子邮件的主题的语言模型。 从输入语音方法生成的文本可以由电子邮件消息处理器来总结，并且可以由用户编辑。 然后可以通过蜂窝连接再次将生成的电子邮件消息发送到Internet电子邮件服务器，以将电子邮件消息发送给接收者。

公开(公告)号：US06571174B2

公开(公告)日：2003-05-27

申请号：US09929634

申请日：2001-08-14

申请人： Luca Rigazio ,  Philippe R. Morin ,  Jean-Claude Junqua

发明人： Luca Rigazio ,  Philippe R. Morin ,  Jean-Claude Junqua

IPC分类号： G01C2134

CPC分类号： G08G1/096811 ,  G01C21/26 ,  G08G1/202

摘要： A navigation apparatus is disclosed which may be used by law enforcement personnel for rapid intervention to a location while adding safety and reliability to the process. The apparatus includes a computer system, having an operating system, memory and a user interface. The system further includes a positioning system, such as a GPS system for determining the position of a vehicle. The positioning system communicates with the operating system. An information database, communicating with the operating system, contains data related to routing information concerning routes for travel by the vehicle. The routing information includes safety information concerning route safety in the traveling region accessible by the vehicle. The apparatus further includes a routing system in communication with the operating system that determines a route based at least in part on the routing information. Driving directions and call information are provided multi-modally to provide the officer with critical information in an efficient and timely fashion.

摘要翻译： 公开了一种导航装置，其可以被执法人员用于对位置的快速干预，同时为该过程增加安全性和可靠性。 该装置包括具有操作系统，存储器和用户界面的计算机系统。 该系统还包括诸如用于确定车辆位置的GPS系统的定位系统。 定位系统与操作系统通信。 与操作系统通信的信息数据库包含与车辆行驶路线有关的路线信息的数据。 路线信息包括关于车辆可接近的行驶区域中的路线安全的安全信息。 该装置还包括与操作系统通信的路由系统，其至少部分地基于路由信息来确定路由。 驾驶方向和通话信息以多方式提供，以有效和及时的方式向官员提供关键信息。