公开(公告)号：US08185480B2

公开(公告)日：2012-05-22

申请号：US12061023

申请日：2008-04-02

申请人： Dimitri Kanevsky ,  David Nahamoo ,  Daniel Povey ,  Bhuvana Ramabhadran

发明人： Dimitri Kanevsky ,  David Nahamoo ,  Daniel Povey ,  Bhuvana Ramabhadran

IPC分类号： G06F15/18

CPC分类号： G06K9/6226

摘要： A method of optimizing a function of a parameter includes associating, with an objective function for initial value of parameters, an auxiliary function of parameters that could be optimized computationally more efficiently than an original objective function, obtaining parameters that are optimum for the auxiliary function, obtaining updated parameters by taking a weighted sum of the optimum of the auxiliary function and initial model parameters.

摘要翻译： 一种优化参数函数的方法包括将参数初始值的目标函数与原始目标函数的计算效率进行优化的参数辅助函数相关联，获得对辅助函数最佳的参数， 通过获取辅助功能和初始模型参数的最优值的加权和获得更新的参数。

公开(公告)号：US08140979B2

公开(公告)日：2012-03-20

申请号：US12192860

申请日：2008-08-15

申请人： Yefim Shuf ,  Alex Zlatsin ,  Dimitri Kanevsky ,  Genady Grabarnik

发明人： Yefim Shuf ,  Alex Zlatsin ,  Dimitri Kanevsky ,  Genady Grabarnik

IPC分类号： G06F3/00 ,  G06F3/048

CPC分类号： G06Q10/06311

摘要： A system and method for managing computer operations according to stored user preferences is disclosed. The system includes a calendar module for scheduling an event and centrally storing the user preferences regarding the computer operations during the event, and a controller module for retrieving the user preferences centrally stored in the calendar module and centrally controlling the computer operations according to the user preferences during the event.

摘要翻译： 公开了一种用于根据存储的用户偏好来管理计算机操作的系统和方法。 系统包括用于调度事件的日历模块，并且在事件期间集中存储关于计算机操作的用户偏好;以及控制器模块，用于检索集中存储在日历模块中的用户偏好，并根据用户偏好集中控制计算机操作 在活动期间。

公开(公告)号：US08103511B2

公开(公告)日：2012-01-24

申请号：US12127874

申请日：2008-05-28

申请人： Sara H. Basson ,  Brian R. Heasman ,  Dimitri Kanevsky ,  Edward Emile Kelley

发明人： Sara H. Basson ,  Brian R. Heasman ,  Dimitri Kanevsky ,  Edward Emile Kelley

IPC分类号： G10L21/00 ,  G10L11/02

CPC分类号： G10L21/02 ,  G11B2020/10546

摘要： An audio file generation method and system. A computing system receives a first audio file comprising first speech data associated with a first party. The computing system receives a second audio file comprising second speech data associated with a second party. The first audio file differs from the second audio file. The computing system generates a third audio file from the second audio file. The third audio file differs from the second audio file. The process to generate the third audio file includes identifying a first set of attributes missing from the second audio file and adding the first set of attributes to the second audio file. The process to generate the third audio file additionally includes removing a second set of attributes from the second audio file. The third audio file includes third speech data associated with the second party. The computing system broadcasts the third audio file.

摘要翻译： 音频文件生成方法和系统。 计算系统接收包括与第一方相关联的第一语音数据的第一音频文件。 计算系统接收包括与第二方相关联的第二语音数据的第二音频文件。 第一个音频文件与第二个音频文件不同。 计算系统从第二音频文件生成第三音频文件。 第三个音频文件与第二个音频文件不同。 生成第三音频文件的过程包括识别从第二音频文件丢失的第一组属性，并将第一组属性添加到第二音频文件。 生成第三音频文件的过程另外包括从第二音频文件中移除第二组属性。 第三音频文件包括与第二方相关联的第三语音数据。 计算系统广播第三个音频文件。

公开(公告)号：US20110191293A1

公开(公告)日：2011-08-04

申请号：US13085106

申请日：2011-04-12

申请人： Dimitri Kanevsky ,  Alexander Zlatsin

发明人： Dimitri Kanevsky ,  Alexander Zlatsin

IPC分类号： G06F17/30

CPC分类号： G06F11/1461 ,  G06F11/1446 ,  G06F11/1456 ,  G06F11/1458 ,  G06F11/1464 ,  G06F11/1471 ,  G06F11/3438 ,  G06F17/30289 ,  G06F2201/80 ,  G06F2201/805 ,  Y10S707/99955

摘要： This invention involves tracking and backing all the information that a user generates on its computer devices (including embedded devices) in real time. The local user server records all user actions and gestures (via various means that include TV cameras). All of this information (user actions and saved files in a computer) is then sent to a remote server via the Internet. This remote server has a virtual map of all the embedded devices on a computer that the person uses. The remote server immediately starts to interpret the user's actions (including user gestures). In one implementation, the invention stores user actions that are related to data generation (e.g. actions that called some links where data is stored, or executed some programs that generated data). In another variant the remote server generates and downloads the same files that are downloaded on the local user computer devices. For example, if a person begins to download a program, the server may also download the same program on a remote backup server. This way, if the user loses this program, it can be retrieved automatically through a provided server on the Internet. If user's files are backed up by regular backup periodically, relevant data that were stored by real time backup servers can be eliminated.

摘要翻译： 本发明涉及跟踪和支持用户在其计算机设备（包括嵌入式设备）上实时产生的所有信息。 本地用户服务器记录所有用户操作和手势（通过包括电视摄像机的各种方法）。 所有这些信息（用户操作和计算机中保存的文件）然后通过Internet发送到远程服务器。 该远程服务器具有人使用的计算机上的所有嵌入式设备的虚拟映射。 远程服务器立即开始解释用户的操作（包括用户手势）。 在一个实现中，本发明存储与数据生成相关的用户动作（例如，称为存储数据的一些链接的动作，或者执行一些生成数据的程序）。 在另一个变体中，远程服务器生成并下载在本地用户计算机设备上下载的相同文件。 例如，如果某人开始下载程序，则服务器还可以在远程备份服务器上下载相同的程序。 这样一来，如果用户丢失了这个程序，就可以通过Internet上提供的服务器自动检索。 如果定期备份用户文件，则可以消除实时备份服务器存储的相关数据。

公开(公告)号：US20110054853A1

公开(公告)日：2011-03-03

申请号：US12551297

申请日：2009-08-31

申请人： Narges Bani Asadi ,  Guillermo A. Cecchi ,  Dimitri Kanevsky ,  Bhuvana Ramabhadran ,  Irina Rish ,  Katya Scheinberg

发明人： Narges Bani Asadi ,  Guillermo A. Cecchi ,  Dimitri Kanevsky ,  Bhuvana Ramabhadran ,  Irina Rish ,  Katya Scheinberg

IPC分类号： G06F17/10

CPC分类号： G06N99/005 ,  G06N7/005

摘要： A method, information processing system, and computer readable article of manufacture model data. A first dataset is received that includes a first set of physical world data. At least one data model associated with the first dataset is generated based on the receiving. A second dataset is received that includes a second set of physical world data. The second dataset is compared to the at least one data model. A probability that the second dataset is modeled by the at least one data model is determined. A determination is made that the probability is above a given threshold. A decision associated with the second dataset based on the at least one data model is generated in response to the probability being above the given threshold. The probability and the decision are stored in memory. The probability and the decision are provided to user via a user interface.

摘要翻译： 一种方法，信息处理系统和计算机可读物品的制造模型数据。 接收包括第一组物理世界数据的第一数据集。 基于接收生成与第一数据集相关联的至少一个数据模型。 接收包括第二组物理世界数据的第二数据集。 将第二数据集与至少一个数据模型进行比较。 确定第二数据集由至少一个数据模型建模的概率。 确定概率高于给定阈值。 响应于高于给定阈值的概率，生成基于至少一个数据模型与第二数据集相关联的决定。 概率和决定存储在内存中。 通过用户界面向用户提供概率和决定。

公开(公告)号：US07864072B2

公开(公告)日：2011-01-04

申请号：US12140543

申请日：2008-06-17

申请人： Sara H. Basson ,  Wilfredo Ferre ,  Julien Ghez ,  Dimitri Kanevsky ,  Frances W. West

发明人： Sara H. Basson ,  Wilfredo Ferre ,  Julien Ghez ,  Dimitri Kanevsky ,  Frances W. West

IPC分类号： G08G1/07

CPC分类号： G08G1/005 ,  G08G1/07

摘要： A system, program product and method for automatically adjusting the traffic light of a traffic light controlled intersection. Personal data relative to a pedestrian cross walking the intersection, including walking speed, and the current speed of a vehicle approaching the intersection are simultaneously acquired. Both the personal data and the vehicle current speed are processed to generate cross walk control signals, such as indicators of risk of collision between vehicle and pedestrian. Where the risk warrants action, the “stop” condition of the traffic light is enable to warn the vehicle to stop. Traffic control signals are also generated to control the duration of the “walk” condition for slow moving pedestrians.

摘要翻译： 一种用于自动调整交通灯控制交叉口交通信号灯的系统，程序产品和方法。 同时获取相对于步行十字路口的行人交叉路口的个人数据，包括步行速度和接近交叉口的车辆的当前速度。 处理个人数据和车辆当前速度以产生交叉步行控制信号，例如车辆和行人之间的碰撞风险的指标。 如果风险值得采取行动，则交通信号灯的“停止”状态能够警告车辆停止。 还产生交通控制信号以控制缓慢移动的行人的“步行”条件的持续时间。

公开(公告)号：US07801727B2

公开(公告)日：2010-09-21

申请号：US11064643

申请日：2005-02-24

申请人： Ponani Gopalakrishnan ,  Dimitri Kanevsky ,  Michael Daniel Monkowski ,  Jan Sedivy

发明人： Ponani Gopalakrishnan ,  Dimitri Kanevsky ,  Michael Daniel Monkowski ,  Jan Sedivy

IPC分类号： G10L15/04

CPC分类号： G10L15/197 ,  G06F17/27 ,  G10L15/183 ,  Y10S707/99942

摘要： A method for generating a language component vocabulary VC for a speech recognition system having a language vocabulary V of a plurality of word forms is disclosed. The method includes: partitioning the language vocabulary V into subsets of word forms based on frequencies of occurrence of the respective word forms; and in at least one of the subsets, splitting word forms having frequencies less than a threshold to thereby generate word form components. Also disclosed is a method for use in speech recognition including: splitting an acoustic vocabulary comprising baseforms into baseform components and storing the baseform components; and, performing sound to spelling mapping on the baseform components so as to generate a baseform components to word parts table for use in subsequent decoding of speech. A method for decoding a speech utterance using language model components and acoustic components, includes the steps of: generating from the utterance a stack of baseform component paths; concatenating baseform components in a path to generate concatenated baseforms, when the concatenated baseform components correspond to a baseform found in an acoustic vocabulary; mapping the concatenated baseforms into words; computing language model (LM) scores associated with the words using a language model, and performing further decoding of the utterance based thereupon.

摘要翻译： 公开了一种用于生成具有多个单词形式的语言词汇V的语音识别系统的语言组件词汇VC的方法。 该方法包括：基于各个词形式的出现频率将语言词汇V划分成单词形式的子集; 并且在至少一个子集中，分割具有小于阈值的频率的字形式，从而生成词形分量。 还公开了一种用于语音识别的方法，包括：将包含基本形式的声学词汇分解成基本形式组件并存储基本形式组件; 并且对基本形式组件执行声音拼写映射，以便生成用于语音后续解码中的字部分表的基本形式分量。 一种使用语言模型分量和声学分量对语音发音进行解码的方法，包括以下步骤：从发音中产生一叠基础分量路径; 当级联的基本形式组件对应于在声学词汇中发现的基础形式时，将路径中的基本形式组件连接以生成级联的基本形式; 将连接的基本形式映射为单词; 与使用语言模型的单词相关联的计算语言模型（LM）得分，并且基于此进行对话语的进一步解码。

公开(公告)号：US07792701B2

公开(公告)日：2010-09-07

申请号：US10779376

申请日：2004-02-13

申请人： Sara H. Basson ,  Peter G. Fairweather ,  Alexander Faisman ,  Dimitri Kanevsky ,  Bradley L. Westpfahl

发明人： Sara H. Basson ,  Peter G. Fairweather ,  Alexander Faisman ,  Dimitri Kanevsky ,  Bradley L. Westpfahl

IPC分类号： G06Q30/00

CPC分类号： H04N7/17336 ,  G06Q30/0601 ,  G06Q30/08 ,  H04N21/234309 ,  H04N21/234336 ,  H04N21/254 ,  H04N21/41407

摘要： An accessibility service tool providing accessibility services on demand for converting between content types and program product therefore. Media input, e.g., audio or audio/video is provided to a transformation and augmentation unit that determines from the media content what type of conversion is required and what service is being requested. An auction manager negotiates with service providers and selects one or more service provider for converting the input content into accessible content that is provided instead of or with the original content. Content conversion may be automatic or manual. Manual conversion may be used in background training for automatic conversion.

摘要翻译： 因此，一个可访问性服务工具提供了在内容类型和程序产品之间转换的可访问性服务。 媒体输入，例如音频或音频/视频被提供给转换和增强单元，其从媒体内容确定需要什么类型的转换以及正在请求什么样的服务。 拍卖经理与服务提供商进行协商，并选择一个或多个服务提供商将输入内容转换成可以提供的可访问内容，而不是原始内容。 内容转换可能是自动或手动的。 手动转换可用于自动转换的后台训练。

公开(公告)号：US07729920B2

公开(公告)日：2010-06-01

申请号：US12103560

申请日：2008-04-15

申请人： Jarir Kamel Chaar ,  Dimitri Kanevsky ,  Roberto Sicconi ,  Mahesh Viswanathan

发明人： Jarir Kamel Chaar ,  Dimitri Kanevsky ,  Roberto Sicconi ,  Mahesh Viswanathan

IPC分类号： G10L15/22

CPC分类号： G10L15/22

摘要： Systems and methods are provided for processing and executing commands in automated systems. For example, command processing systems and methods are provided which can automatically determine, evaluate or otherwise predict consequences of execution of misrecognized or misinterpreted user commands in automated systems and thus prevent undesirable or dangerous consequences that can result from execution of misrecognized/misinterpreted commands.

摘要翻译： 提供了系统和方法来处理和执行自动化系统中的命令。 例如，提供了命令处理系统和方法，其可以自动地确定，评估或以其他方式预测在自动化系统中执行错误识别或误解的用户命令的后果，并且因此防止由于执行错误识别/误解的命令而导致的不期望或危险的后果。

公开(公告)号：US07711105B2

公开(公告)日：2010-05-04

申请号：US12114420

申请日：2008-05-02

申请人： Sara H. Basson ,  Alexander Faisman ,  Dimitri Kanevsky ,  Nitendra Rajput

发明人： Sara H. Basson ,  Alexander Faisman ,  Dimitri Kanevsky ,  Nitendra Rajput

IPC分类号： H04M3/00

CPC分类号： H04M3/5233 ,  H04M2201/40 ,  H04M2201/60 ,  H04M2203/2061

摘要： Techniques for improving the efficiency of foreign accent/language communications processing are disclosed. For example, in one aspect of the invention, a technique for processing a call received by a call center, includes the following steps/operations. A call is obtained. Then, at least one of an accent and a language spoken by a caller making the call is identified. The call is then directed to an appropriate operator at a first level of the call center based on at least one of the automatically identified accent and the automatically identified language. The techniques may also include multiple-purpose editing, automated translation, and restating methodologies.

摘要翻译： 公开了提高外语/语言通信处理效率的技术。 例如，在本发明的一个方面，一种用于处理由呼叫中心接收的呼叫的技术包括以下步骤/操作。 获得通话。 然后，识别由进行呼叫的呼叫者所说的口音和语言中的至少一个。 基于自动识别的口音和自动识别的语言中的至少一个，呼叫被定向到呼叫中心的第一级的适当的操作者。 这些技术还可以包括多用途编辑，自动翻译和重述方法。