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. 发明专利

JP2010179454A Learning and use of schemata in robotic device 审中-公开
标题翻译：在机器人设备中学习和使用SCHEMATA
公开(公告)号：JP2010179454A
公开(公告)日：2010-08-19
申请号：JP2009287375
申请日：2009-12-18
申请人： Honda Research Inst Europe Gmbh , ホンダリサーチインスティテュートヨーロッパゲーエムベーハーHonda Research Institute Europe GmbH
发明人： GLAESER CLAUDIUS , JOUBLIN FRANK
IPC分类号： B25J13/00 , G05D1/02 , G06N3/00 , G06N3/04
CPC分类号： B25J9/161 , G05B2219/33025 , G05B2219/33051 , G05B2219/39285 , G05B2219/39376
摘要： PROBLEM TO BE SOLVED: To provide a method for controlling an autonomous mobile robot with excellent behavior when interacting with environment.
SOLUTION: The method uses schemata being a group of sequences of motion commands expressed by parameters to make a robot achieve a set goal. The parameters for the sequences are gained from the state variables of a robotic controller. The robotic controller includes: an interface for supplying a sensory input to the robotic controller; a schema state memory supplied with either the input from a schema recognition module, the input from an inverse model module or a combination of them; an inverse model for generating a motion command based on the state variables and stored schemata; a forward model for predicting the state variables based on the state variables and stored schemata; and the schema recognition module for selecting a schema based on the supplied state variables of the robot controlled by the robotic controller.
COPYRIGHT: (C)2010,JPO&INPIT
摘要翻译：要解决的问题：提供一种用于在与环境相互作用时具有优异行为的自动移动机器人的控制方法。
解决方案：该方法使用由参数表达的运动命令序列组成的模式，以使机器人达到设定目标。序列的参数从机器人控制器的状态变量获得。机器人控制器包括：用于向机器人控制器提供感觉输入的接口; 模式状态存储器，其提供有来自模式识别模块的输入，来自逆模型模块的输入或它们的组合; 基于状态变量和存储模式生成运动指令的逆模型; 一种基于状态变量和存储模式预测状态变量的前向模型; 以及所述模式识别模块，用于基于由所述机器人控制器控制的所述机器人提供的状态变量来选择模式。版权所有（C）2010，JPO＆INPIT

2. 发明专利

JP2008090295A Joint estimation of formant trajectory via bayesian technique and adaptive segmentation 有权
标题翻译：通过BAYESIAN技术和自适应分类的联合估计
公开(公告)号：JP2008090295A
公开(公告)日：2008-04-17
申请号：JP2007231886
申请日：2007-09-06
申请人： Honda Research Inst Europe Gmbh , ホンダリサーチインスティテュートヨーロッパゲーエムベーハーHonda Research Institute Europe GmbH
发明人： JOUBLIN FRANK , HECKMANN MARTIN , GLAESER CLAUDIUS
IPC分类号： G10L15/02 , G10L25/15 , G10L25/48
CPC分类号： G10L25/48 , G10L25/15
摘要： PROBLEM TO BE SOLVED: To provide the invention which relates to the field of automated processing of speech signals and particularly to a method for tracking the formant frequencies in a speech signal.
SOLUTION: The method comprises the steps of: obtaining an auditory image of the speech signal; sequentially estimating formant locations; segmenting the frequency range into sub-regions; smoothing the obtained component filtering distributions; and calculating exact formant locations.
COPYRIGHT: (C)2008,JPO&INPIT
摘要翻译：要解决的问题：提供涉及语音信号的自动处理领域的本发明，特别是涉及用于跟踪语音信号中的共振峰频率的方法。解决方案：该方法包括以下步骤：获得语音信号的听觉图像; 顺序地估计共振峰位置; 将频率范围分割为子区域; 平滑所获得的分量滤波分布; 并计算确切的共振峰位置。版权所有（C）2008，JPO＆INPIT

3. 发明专利

JP2007322416A On-line compensation method for converting auditory sense cue for inferring location of sound source into location 审中-公开
标题翻译：用于转换用于感知声源位置到位置的审核感知的线性补偿方法
公开(公告)号：JP2007322416A
公开(公告)日：2007-12-13
申请号：JP2007078729
申请日：2007-03-26
申请人： Honda Research Inst Europe Gmbh , ホンダリサーチインスティテュートヨーロッパゲーエムベーハーHonda Research Institute Europe GmbH
发明人： JOUBLIN FRANK , RODEMANN TOBIAS
IPC分类号： G01S3/808 , B25J13/08 , H04R3/00 , H04R5/027
CPC分类号： G01S3/8055 , G01S3/8083
摘要： PROBLEM TO BE SOLVED: To provide method for adapting sound source localization system on-line, that is, a localized operating method.
SOLUTION: This is the method comprises a step (210) to direct first and second microphones toward the sound source, a step (220) to measure both ears time difference (ITD) between the two microphones, and a step (230) of repeating the two steps (210, 220), until the both-ear time difference (ITD) becomes smaller than the predefined threshold (ε). For a pair of the first and second microphones, it is desirable that this method additionally include a step of directing third microphone toward the sound source, a step of measuring the both-ear time difference (ITD) between the second microphone and the third microphone, and a step of repeating the two steps, until the both-ear time difference (ITD) becomes smaller than a predefined threshold (ε
2 ).
COPYRIGHT: (C)2008,JPO&INPIT
摘要翻译：要解决的问题：提供在线调整声源定位系统的方法，即本地化操作方法。解决方案：该方法包括将第一和第二麦克风引向声源的步骤（210），测量两个麦克风之间的耳朵时间差（ITD）的步骤（220），以及步骤（230））重复两个步骤（210,220），直到双耳时间差（ITD）变得小于预定阈值（ε）。对于一对第一和第二麦克风，期望该方法还包括将第三麦克风引向声源的步骤，测量第二麦克风和第三麦克风之间的双耳时间差（ITD）的步骤以及重复两个步骤的步骤，直到双耳时间差（ITD）变得小于预定阈值（εSB> 2 ）。版权所有（C）2008，JPO＆INPIT

4. 发明专利

JP2005346079A Unified processing method of resolved harmonic and unresolved harmonic 有权
标题翻译：分解谐波和非对称谐波的统一处理方法
公开(公告)号：JP2005346079A
公开(公告)日：2005-12-15
申请号：JP2005162484
申请日：2005-06-02
申请人： Honda Research Inst Europe Gmbh , ホンダリサーチインスティテュートヨーロッパゲーエムベーハーHonda Research Institute Europe GmbH
发明人： JOUBLIN FRANK , HECKMANN MARTIN
IPC分类号： G10L21/0272 , G10L25/18 , G10L21/02
CPC分类号： G10L21/0272 , G10L25/18
摘要： PROBLEM TO BE SOLVED: To provide a method for processing a resolved harmonic and an unresolved harmonic using a same algorithm and combining the result.
SOLUTION: The method provided by this invention is to subsequently process a modulation envelope obtained from a result of mutual interferences between two harmonics in one filter zone, and includes a step for filtering the modulation envelope through a bandpass filter bank. A combination of demodulation to the modulation envelope and application of the bandpass filters enables applying the same algorithm to the resolved harmonic and the unresolved harmonic.
COPYRIGHT: (C)2006,JPO&NCIPI
摘要翻译：要解决的问题：提供一种使用相同算法处理分辨谐波和未解决谐波的方法，并结合结果。解决方案：本发明提供的方法是随后处理从一个滤波器区域中的两个谐波之间的相互干扰的结果获得的调制包络，并且包括通过带通滤波器组对调制包络进行滤波的步骤。解调到调制包络和带通滤波器的应用的组合使得能够将相同的算法应用于解析谐波和未解决的谐波。版权所有（C）2006，JPO＆NCIPI

5. 发明专利

JP2008005472A Robot head with artificial ear 有权
标题翻译：机器人与人工耳朵
公开(公告)号：JP2008005472A
公开(公告)日：2008-01-10
申请号：JP2007125763
申请日：2007-05-10
申请人： Honda Research Inst Europe Gmbh , ホンダリサーチインスティテュートヨーロッパゲーエムベーハーHonda Research Institute Europe GmbH
发明人： JOUBLIN FRANK
IPC分类号： H04R5/027 , B25J19/02
CPC分类号： B25J13/003 , B25J19/026
摘要： PROBLEM TO BE SOLVED: To provide a robot head with artificial ears.
SOLUTION: A supporting head is provided with binaural artificial ears. Each binaural artificial ear comprises an auricle-shaped structure and a microphone. In order to enhance auditory localization cues, at least an upper part of the head is provided with an acoustically dampening surface.
COPYRIGHT: (C)2008,JPO&INPIT
摘要翻译：要解决的问题：为机器人头部提供人造耳朵。
解决方案：支撑头设有双耳人造耳。每个双耳人造耳包括耳廓形结构和麦克风。为了增强听觉定位线索，头部的至少上部设置有声学抑制表面。版权所有（C）2008，JPO＆INPIT

6. 发明专利

JP2006338660A Method for developing parallel/distributed application 审中-公开
标题翻译：用于开发并行/分布式应用的方法
公开(公告)号：JP2006338660A
公开(公告)日：2006-12-14
申请号：JP2006142670
申请日：2006-05-23
申请人： Honda Research Inst Europe Gmbh , ホンダリサーチインスティテュートヨーロッパゲーエムベーハーHonda Research Institute Europe GmbH
发明人： JOUBLIN FRANK , GOERICK CHRISTIAN , CERAVOLA ANTONELLO , DUNN MARK
IPC分类号： G06F9/44
CPC分类号： G06F11/3664 , G06F8/34
摘要： PROBLEM TO BE SOLVED: To provide improved support measures for designing, making, and testing of a parallel/distributed application. SOLUTION: The method is used for developing the parallel/distributed application. A process in the development includes a designing stage, an implementing stage, and a testing stage. The method is composed so as to include a step for providing a script language for expressing elements in a connection graph and connection between the elements in the designing stage, a step for providing the previously-defined module to implement functions of the application in the implementing stage, a step for providing the previously-defined executor to define a type of execution of the module in the implementing stage, and a step for providing the previously-defined process instance to distribute the application on a plurality of arithmetic devices in the implementing stage, and a step for providing the previously-defined abstraction level to monitor and test the application under testing in the testing stage. COPYRIGHT: (C)2007,JPO&INPIT
摘要翻译：要解决的问题：为并行/分布式应用程序的设计，制作和测试提供改进的支持措施。解决方案：该方法用于开发并行/分布式应用程序。开发过程包括设计阶段，实施阶段和测试阶段。该方法被构成为包括提供用于在连接图中表达元素的脚本语言和在设计阶段中的元素之间的连接的步骤，用于提供先前定义的模块以在实现中实现应用的功能的步骤阶段，提供先前定义的执行器以在实现阶段中定义模块的执行类型的步骤，以及用于提供先前定义的流程实例以在实现阶段中的多个算术设备上分发应用的步骤，以及提供先前定义的抽象级别以在测试阶段监测和测试应用程序的步骤。版权所有（C）2007，JPO＆INPIT

7. 发明专利

JP2006209123A Method of finding fundamental frequency of higher harmonic signal 有权
标题翻译：发现更高谐波信号的基本频率的方法
公开(公告)号：JP2006209123A
公开(公告)日：2006-08-10
申请号：JP2006015950
申请日：2006-01-25
申请人： Honda Research Inst Europe Gmbh , ホンダリサーチインスティテュートヨーロッパゲーエムベーハーHonda Research Institute Europe GmbH
发明人： JOUBLIN FRANK , HECKMANN MARTIN
IPC分类号： G10L25/90
CPC分类号： G10L25/90
摘要： PROBLEM TO BE SOLVED: To provide a new method of finding a common fundamental frequency of higher harmonics of a higher harmonic signal.
SOLUTION: Provided is the method of finding the fundamental frequency of the higher harmonic signal. This method includes a step of dividing the higher harmonic signal into different frequency channels by a band-pass filter, a step of calculating distances of zero crossings of different degrees by the frequency channels, and a step of calculating a histogram of all distance values at respective moments of time. Here, the distance values have a peak at a distance corresponding to the fundamental frequency.
COPYRIGHT: (C)2006,JPO&NCIPI
摘要翻译：要解决的问题：提供一种找到高次谐波信号的高次谐波的共同基频的新方法。解决方案：提供了找到高次谐波信号的基频的方法。该方法包括通过带通滤波器将高次谐波信号划分成不同频率信道的步骤，通过频率信道计算不同程度的零交叉点的距离的步骤，以及计算所有距离值的直方图的步骤各自的时刻。这里，距离值在与基本频率对应的距离处具有峰值。版权所有（C）2006，JPO＆NCIPI

8. 发明专利

JP2008242431A Pitch extraction with inhibition of harmonic and sub-harmonic of fundamental frequency 有权
标题翻译：具有抑制基本频率的谐波和次谐波的PITCH提取
公开(公告)号：JP2008242431A
公开(公告)日：2008-10-09
申请号：JP2008013165
申请日：2008-01-23
申请人： Honda Research Inst Europe Gmbh , ホンダリサーチインスティテュートヨーロッパゲーエムベーハーHonda Research Institute Europe GmbH
发明人： JOUBLIN FRANK , HECKMANN MARTIN
IPC分类号： G10L25/90
CPC分类号： G10L25/90 , G10L25/18
摘要： PROBLEM TO BE SOLVED: To provide a robust method for estimating the fundamental frequency of a harmonic signal.
SOLUTION: The method for estimating the fundamental frequency of the harmonic signal comprises the steps: - forming a fundamental frequency hypothesis (f0'); - providing a comb filter based on the fundamental frequency hypothesis; - filtering the given harmonic signal using the comb filter; and - testing the fundamental frequency hypothesis for each tooth in the comb filter.
COPYRIGHT: (C)2009,JPO&INPIT
摘要翻译：要解决的问题：提供用于估计谐波信号的基频的鲁棒方法。解决方案：用于估计谐波信号的基频的方法包括以下步骤： - 形成基频假设（f0'）; - 提供基于频率假设的梳状滤波器; - 使用梳状滤波器对给定谐波信号进行滤波; 并测试梳状滤波器中每个齿的基频估计。版权所有（C）2009，JPO＆INPIT

9. 发明专利

JP2007044865A Orientation/positioning system control method 有权
标题翻译：定位/定位系统控制方法
公开(公告)号：JP2007044865A
公开(公告)日：2007-02-22
申请号：JP2006210060
申请日：2006-08-01
申请人： Honda Research Inst Europe Gmbh , ホンダリサーチインスティテュートヨーロッパゲーエムベーハーHonda Research Institute Europe GmbH
发明人： JOUBLIN FRANK , RODEMANN TOBIAS
IPC分类号： B25J13/08 , G02B7/28 , G03B15/00
CPC分类号： B25J9/163 , B25J9/1697 , G05B19/401 , G05B2219/39367 , G05B2219/39387 , G05B2219/40617
摘要： PROBLEM TO BE SOLVED: To provide an orientation system control method capable of automatically calibrating a system highly in precision.
SOLUTION: This method includes a step S1 to evaluate before-behavior output information of a sensor means to detect a pattern position in an input space of the sensor means, a step S2 to decide an after-behavior position made a target of the pattern in the input space of the sensor means, a step S3 to define a command concerning an actuator means by mapping deflection of the before-behavior position and the target after-behavior position on an input space coordinate of the sensor means on an actuator control coordinate by using a specified mapping function, a step S4 to carry out orientation/positioning of the sensor means by the actuator means in accordance with the defined command to execute orientation/positioning behavior, a step S5 to detect the actual after-behavior position of the pattern in the input space of the sensor means, a step S6 to detect outside motion of the pattern and a step 7 suitable for the mapping function used in the step S3.
COPYRIGHT: (C)2007,JPO&INPIT
摘要翻译：要解决的问题：提供一种能够精确地自动校准系统的定向系统控制方法。解决方案：该方法包括步骤S1，用于评估用于检测传感器装置的输入空间中的图案位置的传感器装置的行为前输出信息;步骤S2，用于确定制成目标的行为后位置在传感器装置的输入空间中的图案，通过将传感器装置的输入空间坐标上的行为位置和目标后行为位置的偏转映射在致动器装置的致动器装置的输入空间坐标上来限定关于致动器装置的命令的步骤S3 通过使用指定的映射功能的控制坐标;步骤S4，根据所定义的命令执行传感器装置的定向/定位，以执行定向/定位行为;步骤S5，用于检测实际的行为后位置传感器装置的输入空间中的图案的步骤S6，用于检测图案的外部运动的步骤S6和适于在步骤S3中使用的映射功能的步骤7。版权所有（C）2007，JPO＆INPIT

10. 发明专利

JP2008090272A Using child directed speech to bootstrap model based speech segmentation and recognition system 有权
标题翻译：使用儿童指导语音到基于模式的语音分类和识别系统
公开(公告)号：JP2008090272A
公开(公告)日：2008-04-17
申请号：JP2007148155
申请日：2007-06-04
申请人： Honda Research Inst Europe Gmbh , ホンダリサーチインスティテュートヨーロッパゲーエムベーハーHonda Research Institute Europe GmbH
发明人： JOUBLIN FRANK , BRANDL HOLGER
IPC分类号： G10L15/06 , G10L15/04
CPC分类号： G10L15/04 , G10L15/063
摘要： PROBLEM TO BE SOLVED: To provide an improved method and system for automation for acquiring an unsupervised acoustic model.
SOLUTION: The method includes steps of: finding a training segment (X) using unsupervised speech segmentation or SU spotting; if a model pool (P) is empty, training a first speech syllable model and adding it to the pool; else determining an existing model (F) in the model pool (P) that best matches the given segment (X); using the model (F) to score the new training segment (X); if the score is smaller than a predefined threshold, creating a new model for the training segment (X) and adding it to the pool; else using the new training segment (X) to improve or to re-estimate the model (F).
COPYRIGHT: (C)2008,JPO&INPIT
摘要翻译：要解决的问题：提供用于获取无监督声学模型的自动化的改进的方法和系统。解决方案：该方法包括以下步骤：使用无监督的语音分割或SU识别找到训练段（X）; 如果模型池（P）为空，则训练第一个语音音节模型并将其添加到池中; 否则确定与给定段（X）最佳匹配的模型池（P）中的现有模型（F）; 使用模型（F）对新的训练段（X）进行评分; 如果分数小于预定阈值，则为训练段（X）创建新模型并将其添加到池中; 否则使用新的训练段（X）来改进或重新估计模型（F）。版权所有（C）2008，JPO＆INPIT

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