专利快速检索-快速检索全球专利，免费商用专利数据库-IPRDB

51. 发明申请

US20110288867A1 NAMETAG CONFUSABILITY DETERMINATION 有权
标题翻译： NAMETAG可信度测定
公开(公告)号：US20110288867A1
公开(公告)日：2011-11-24
申请号：US12782141
申请日：2010-05-18
申请人： Rathinavelu Chengalvarayan , Lawrence D. Cepuran
发明人： Rathinavelu Chengalvarayan , Lawrence D. Cepuran
IPC分类号： G10L15/04
CPC分类号： G10L15/1815 , G10L13/08 , G10L15/063 , G10L15/187 , G10L2015/0631
摘要： A method of and system for managing nametags including receiving a command from a user to store a nametag, prompting the user to input a number to be stored in association with the nametag, receiving an input for the number from the user, prompting the user to input the nametag to be stored in association with the number, receiving an input for the nametag from the user, processing the nametag input, and calculating confusability of the nametag input in multiple individual domains including a nametag domain, a number domain, and a command domain.
摘要翻译：一种用于管理名称的方法和系统，包括从用户接收命令以存储名称，提示用户输入要与该名称相关联地存储的号码，从用户接收该号码的输入，提示用户输入要与该号码相关联地存储的名称，从用户接收用于命名的输入，处理命名输入，以及计算在多个单个域中的名称输入的混淆，包括命名域，数字域和命令域。

52. 发明授权

US07983916B2 Sampling rate independent speech recognition 有权
标题翻译：采样率独立语音识别
公开(公告)号：US07983916B2
公开(公告)日：2011-07-19
申请号：US11772992
申请日：2007-07-03
申请人： Rathinavelu Chengalvarayan
发明人： Rathinavelu Chengalvarayan
IPC分类号： G10L15/02 , G10L15/06
CPC分类号： G10L15/02
摘要： A sampling-rate-independent method of automated speech recognition (ASR). Speech energies of a plurality of codebooks generated from training data created at an ASR sampling rate are compared to speech energies in a current frame of acoustic data generated from received audio created at an audio sampling rate below the ASR sampling rate. A codebook is selected from the plurality of codebooks, and has speech energies that correspond to speech energies in the current frame over a spectral range corresponding to the audio sampling rate. Speech energies above the spectral range are copied from the selected codebook and appended to the current frame.
摘要翻译：自动语音识别（ASR）的采样率独立方法。将从以ASR采样率创建的训练数据生成的多个码本的语音能量与在以低于ASR采样率的音频采样率产生的接收音频生成的声音数据的当前帧中的语音能量进行比较。从多个码本中选择码本，并且具有对应于当前帧中对应于音频采样率的频谱范围上的语音能量的语音能量。高于光谱范围的语音能量从所选码本中复制并附加到当前帧。

53. 发明授权

US07676363B2 Automated speech recognition using normalized in-vehicle speech 有权
标题翻译：使用归一化车载语音的自动语音识别
公开(公告)号：US07676363B2
公开(公告)日：2010-03-09
申请号：US11427590
申请日：2006-06-29
申请人： Rathinavelu Chengalvarayan , Scott M Pennock
发明人： Rathinavelu Chengalvarayan , Scott M Pennock
IPC分类号： G10L15/20 , G10L19/14 , G10L21/00
CPC分类号： G10L15/20
摘要： A speech recognition method includes the steps of receiving speech in a vehicle, extracting acoustic data from the received speech, and applying a vehicle-specific inverse impulse response function to the extracted acoustic data to produce normalized acoustic data. The speech recognition method may also include one or more of the following steps: pre-processing the normalized acoustic data to extract acoustic feature vectors; decoding the normalized acoustic feature vectors using as input at least one of a plurality of global acoustic models built according to a plurality of Lombard levels of a Lombard speech corpus covering a plurality of vehicles; calculating the Lombard level of vehicle noise; and/or selecting the at least one of the plurality of global acoustic models that corresponds to the calculated Lombard level for application during the decoding step.
摘要翻译：语音识别方法包括以下步骤：在车辆中接收语音，从接收的语音中提取声音数据，以及将车辆特定的反向脉冲响应函数应用于所提取的声学数据，以产生归一化的声学数据。语音识别方法还可以包括以下一个或多个步骤：预处理归一化声学数据以提取声学特征向量; 使用根据覆盖多个车辆的Lombard语音语料库的多个Lombard级别构建的多个全局声学模型中的至少一个，对归一化的声学特征向量进行解码; 计算车辆噪声的隆巴德水平; 和/或选择与在所述解码步骤期间应用的所计算的伦巴第级别对应的所述多个全局声学模型中的至少一个。

54. 发明申请

US20080126091A1 VOICE DIALING USING A REJECTION REFERENCE 有权
标题翻译：使用拒绝参考的语音拨号
公开(公告)号：US20080126091A1
公开(公告)日：2008-05-29
申请号：US11563809
申请日：2006-11-28
申请人： Jason W. Clark , Rathinavelu Chengalvarayan , Timothy J. Grost , Dana B. Fecher , Jeremy M. Spaulding
发明人： Jason W. Clark , Rathinavelu Chengalvarayan , Timothy J. Grost , Dana B. Fecher , Jeremy M. Spaulding
IPC分类号： G10L17/00
CPC分类号： G10L15/22 , G10L2015/0631 , H04M1/271 , H04M1/6075
摘要： A voice dialing method includes the steps of receiving an utterance from a user, decoding the utterance to identify a recognition result for the utterance, and communicating to the user the recognition result. If an indication is received from the user that the communicated recognition result is incorrect, then it is added to a rejection reference. Then, when the user repeats the misunderstood utterance, the rejection reference can be used to eliminate the incorrect recognition result as a potential subsequent recognition result. The method can be used for single or multiple digits or digit strings.
摘要翻译：语音拨号方法包括以下步骤：从用户接收话语，解码话语以识别语音的识别结果，以及向用户传达识别结果。如果从用户接收到所传送的识别结果不正确的指示，则将其添加到拒绝参考。然后，当用户重复误解话语时，可以使用拒绝参考来消除不正确的识别结果作为潜在的后续识别结果。该方法可用于单个或多个数字或数字串。

55. 发明授权

US06889189B2 Speech recognizer performance in car and home applications utilizing novel multiple microphone configurations 有权
标题翻译：使用新型多麦克风配置的汽车和家庭应用中的语音识别器性能
公开(公告)号：US06889189B2
公开(公告)日：2005-05-03
申请号：US10672167
申请日：2003-09-26
申请人： Robert Boman , Luca Rigazio , Brian Hanson , Rathinavelu Chengalvarayan
发明人： Robert Boman , Luca Rigazio , Brian Hanson , Rathinavelu Chengalvarayan
IPC分类号： G10L15/20 , G10L21/02 , G10L21/00
CPC分类号： G10L21/0208 , G10L15/20 , G10L2021/02166
摘要： System speakers are switched to function as sound input transducers to improve recognizer performance and to support recognizer features. A crossbar switch is selectively activated, either manually or under software control, to allow system loudspeakers to function as sound input transducers that supplement the recognition system microphone or microphone array. Using loudspeakers as “microphones” improves speech recognition in noisy environments, thus attaining better recognition performance with little added system cost. The loudspeakers, positioned in physically separate locations also provide spatial information that can be used to determine the location of the person speaking and thereby offer different functionality for different persons. Acoustic models are selected based on environmental and vehicle operating conditions and may be adapted dynamically using ambient information obtained using the loudspeakers as sound input transducers.
摘要翻译：系统扬声器切换为声音输入传感器，以提高识别器性能并支持识别器功能。手动或软件控制下有选择地激活交叉开关，以允许系统扬声器作为补充识别系统麦克风或麦克风阵列的声音输入换能器。使用扬声器作为“麦克风”可以改善嘈杂环境中的语音识别，从而获得更好的识别性能，增加系统成本。放置在物理上分开的位置的扬声器还提供空间信息，其可用于确定说话者的位置，从而为不同的人提供不同的功能。基于环境和车辆操作条件选择声学模型，并且可以使用使用扬声器获得的环境信息作为声音输入换能器动态地进行调整。

56. 发明授权

US6076058A Linear trajectory models incorporating preprocessing parameters for speech recognition 失效
标题翻译：包含用于语音识别的预处理参数的线性轨迹模型
公开(公告)号：US6076058A
公开(公告)日：2000-06-13
申请号：US32900
申请日：1998-03-02
申请人： Rathinavelu Chengalvarayan
发明人： Rathinavelu Chengalvarayan
IPC分类号： G10L15/065 , G10L15/10 , G10L25/24 , G10L25/27 , G10L15/14
CPC分类号： G10L15/065 , G10L15/10 , G10L25/24 , G10L25/27
摘要： The proposed model aims at finding an optimal linear transformation on the Mel-warped DFT features according to the minimum classification error (MCE) criterion. This linear transformation, along with the (NSHMM) parameters, are automatically trained using the gradient descent method. An advantageous error rate reduction can be realized on a standard 39-class TIMIT phone classification task in comparison with the MCE-trained NSHMM using conventional preprocessing techniques.
摘要翻译：所提出的模型旨在根据最小分类误差（MCE）准则找到Mel翘曲DFT特征的最优线性变换。这种线性变换与（NSHMM）参数一起使用梯度下降法自动训练。与使用常规预处理技术的MCE训练的NSHMM相比，可以在标准的39级TIMIT电话分类任务上实现有利的差错率降低。

57. 发明授权

US6055499A Use of periodicity and jitter for automatic speech recognition 失效
标题翻译：使用周期性和抖动进行自动语音识别
公开(公告)号：US6055499A
公开(公告)日：2000-04-25
申请号：US71214
申请日：1998-05-01
申请人： Rathinavelu Chengalvarayan , David Lynn Thomson
发明人： Rathinavelu Chengalvarayan , David Lynn Thomson
IPC分类号： G10L15/02
CPC分类号： G10L15/02
摘要： A class of features related to voicing parameters that indicate whether the vocal chords are vibrating. Features describing voicing characteristics of speech signals are integrated with an existing 38-dimensional feature vector consisting of first and second order time derivatives of the frame energy and of the cepstral coefficients with their first and second derivatives. Hidden Markov Model (HMM)-based connected digit recognition experiments comparing the traditional and extended feature sets show that voicing features and spectral information are complementary and that improved speech recognition performance is obtained by combining the two sources of information.
摘要翻译：一组与发声参数相关的特征，指示声带是否振动。描述语音信号的发音特征的特征与由帧能量和倒频谱系数的第一和第二阶导数以及它们的第一和第二导数组成的现有38维特征向量集成。基于隐马尔可夫模型（HMM）的连接数字识别实验比较了传统和扩展特征集，表明发声特征和光谱信息是互补的，通过组合两个信息源获得改进的语音识别性能。

58. 发明授权

US08688451B2 Distinguishing out-of-vocabulary speech from in-vocabulary speech 有权
标题翻译：将词汇外的词汇与词汇表达式区分开来
公开(公告)号：US08688451B2
公开(公告)日：2014-04-01
申请号：US11382789
申请日：2006-05-11
申请人： Timothy J. Grost , Rathinavelu Chengalvarayan
发明人： Timothy J. Grost , Rathinavelu Chengalvarayan
IPC分类号： G10L15/00 , G10L15/18 , G10L21/00 , H04M1/64
CPC分类号： G10L15/32
摘要： A speech recognition method includes receiving input speech from a user, processing the input speech using a first grammar to obtain parameter values of a first N-best list of vocabulary, comparing a parameter value of a top result of the first N-best list to a threshold value, and if the compared parameter value is below the threshold value, then additionally processing the input speech using a second grammar to obtain parameter values of a second N-best list of vocabulary. Other preferred steps include: determining the input speech to be in-vocabulary if any of the results of the first N-best list is also present within the second N-best list, but out-of-vocabulary if none of the results of the first N-best list is within the second N-best list; and providing audible feedback to the user if the input speech is determined to be out-of-vocabulary.
摘要翻译：一种语音识别方法，包括从用户接收输入语音，使用第一语法处理输入语音，以获得第一N最佳词汇列表的参数值，将第一N最佳列表的最高结果的参数值与阈值，并且如果比较的参数值低于阈值，则使用第二语法另外处理输入语音，以获得词汇表的第二N最佳列表的参数值。其他优选步骤包括：如果第N个最佳列表的任何结果也存在于第二N最佳列表内，则将输入语音确定为词汇表，但是如果没有结果第一个N最佳列表在第二个N最佳列表中; 以及如果所述输入语音被确定为超出词汇量，则向用户提供可听见的反馈。

59. 发明授权

US08054990B2 Method of recognizing speech from a plurality of speaking locations within a vehicle 有权
标题翻译：从车辆内的多个说话位置识别语音的方法
公开(公告)号：US08054990B2
公开(公告)日：2011-11-08
申请号：US11562853
申请日：2006-11-22
申请人： Jesse T. Gratke , Rathinavelu Chengalvarayan
发明人： Jesse T. Gratke , Rathinavelu Chengalvarayan
IPC分类号： H04R3/00
CPC分类号： B60R16/0373
摘要： A speech recognition method includes the steps of receiving a location-specific command from a vehicle occupant, and adjusting either the shape or magnitude of a pick up pattern of at least one microphone in response to the location-specific command. The microphone adjustment can be carried out by electronically or physically steering the pick-up pattern.
摘要翻译：语音识别方法包括以下步骤：响应于位置特定命令，从车辆乘客接收位置特定命令，以及调整至少一个麦克风的拾取图案的形状或幅度。麦克风调节可以通过电子地或物理地转向拾音图案来进行。

60. 发明授权

US07729911B2 Speech recognition method and system 有权
标题翻译：语音识别方法和系统
公开(公告)号：US07729911B2
公开(公告)日：2010-06-01
申请号：US11235961
申请日：2005-09-27
申请人： Rathinavelu Chengalvarayan , Scott M. Pennock
发明人： Rathinavelu Chengalvarayan , Scott M. Pennock
IPC分类号： G10L15/06
CPC分类号： G10L15/20 , G10L2015/228 , G10L2021/03646
摘要： A speech recognition method comprising the steps of: storing multiple recognition models for a vocabulary set, each model distinguished from the other models in response to a Lombard characteristic, detecting at least one speaker utterance in a motor vehicle, selecting one of the multiple recognition models in response to a Lombard characteristic of the at least one speaker utterance, utilizing the selected recognition model to recognize the at least one speaker utterance; and providing a signal in response to the recognition.
摘要翻译：一种语音识别方法，包括以下步骤：存储用于词汇集的多个识别模型，每个模型响应于伦巴第（Lombard）特征与其他模型区分开来，检测机动车辆中的至少一个说话人话语，选择多个识别模型之一响应于所述至少一个扬声器话语的伦巴第特性，利用所选择的识别模型识别所述至少一个扬声器话语; 并提供响应于识别的信号。

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