专利快速检索-快速检索全球专利，免费商用专利数据库-IPRDB

41. 发明专利

JP2006287810A Contact type microphone 审中-公开
标题翻译：接触式麦克风
公开(公告)号：JP2006287810A
公开(公告)日：2006-10-19
申请号：JP2005107879
申请日：2005-04-04
申请人： Nissan Motor Co Ltd , 日産自動車株式会社
发明人： KAMINUMA MITSUNOBU
IPC分类号： H04R1/46
摘要： PROBLEM TO BE SOLVED: To provide a contact type microphone capable of accurately collecting human speaking voices.
SOLUTION: The contact type microphone which is used in contact with a human body and comprises a microphone unit 1 which collects intra-body conduction sound of human speaking voices and converts them into electric signals, comprises a pressure regulation section 2 for regulating pressure to be applied to the microphone section 1 when the microphone section 1 is in contact with the human body.
COPYRIGHT: (C)2007,JPO&INPIT
摘要翻译：要解决的问题：提供能够准确地收集人声的接触式麦克风。解决方案：与人体接触使用的接触型麦克风包括：麦克风单元1，其收集人体声音的体内传导声音并将其转换为电信号，包括用于调节的压力调节部分2 当麦克风部分1与人体接触时施加到麦克风部分1的压力。版权所有（C）2007，JPO＆INPIT

42. 发明专利

JP2006084898A Sound input device 审中-公开
标题翻译：声音输入设备
公开(公告)号：JP2006084898A
公开(公告)日：2006-03-30
申请号：JP2004270772
申请日：2004-09-17
申请人： Nissan Motor Co Ltd , 日産自動車株式会社
发明人： KAMINUMA MITSUNOBU
IPC分类号： G10L15/20 , G10L21/02
摘要： PROBLEM TO BE SOLVED: To provide a sound input device capable of removing diffusive noise with a slight increase in calculation quantity as compared with SBE applied to a general frequency domain ICA.
SOLUTION: The sound input device detects a sound including a target sound and a non-target sound by a plurality of microphones 10-1 to 10-n to obtain a plurality of sound signals including target sound signals and non-target sound signals and implements an independent component analyzing method of acquiring, by repetitive learning, a filter separating one target sound signal from the sound signals. A sound frequency band is divided into a plurality of frequency division bands at a band division stage 30, discrimination levels for evaluating target sound separation performance of the filter obtained by the learning are calculated by the frequency division bands at an evaluation stage 50, and a frequency division band for learning calculation is determined at a determination stage 60 among the frequency division bands based upon the discrimination levels.
COPYRIGHT: (C)2006,JPO&NCIPI
摘要翻译：要解决的问题：提供一种与应用于一般频域ICA的SBE相比，能够轻微地增加计算量来消除扩散噪声的声音输入装置。解决方案：声音输入装置通过多个麦克风10-1至10-n检测包括目标声音和非目标声音的声音，以获得包括目标声音信号和非目标声音的多个声音信号信号和实现独立的分量分析方法，通过重复学习获得将一个目标声音信号与声音信号分离的滤波器。在频带分割级30，将声频带分割为多个分频频带，通过评价阶段50的分频频带来计算用于评估通过学习获得的滤波器的目标声音分离性能的判别水平，基于鉴别水平，在分频频带中的判定级60确定用于学习计算的分频频带。版权所有（C）2006，JPO＆NCIPI

43. 发明专利

JP2005189289A Speech recognition device, and dictionary preparation method for speech recognition device 有权
标题翻译：语音识别装置和用于语音识别装置的词典准备方法
公开(公告)号：JP2005189289A
公开(公告)日：2005-07-14
申请号：JP2003427331
申请日：2003-12-24
申请人： Nissan Motor Co Ltd , 日産自動車株式会社
发明人： KAMINUMA MITSUNOBU , RI AKINOBU
IPC分类号： G10L15/18 , G10L15/00 , G10L15/06 , G10L15/28
摘要： PROBLEM TO BE SOLVED: To provide a speech recognition device capable of being flexibly compatible with application by effectively using memory, for the reason that although a conventional speech recognition device has utilized a network language dictionary and a statistical language dictionary, any of them has both advantages and disadvantages, and a practical performance has not been achieved, and moreover, although a method of combining them has also been offered, a sufficient result has not been obtained yet.
SOLUTION: The speech recognition device uses a network syntax language dictionary and the statistical language dictionary together, and extracts a vocabulary having much frequency information to be connected to a vocabulary A included in the former language dictionary from the latter language dictionary, and adds this to the former language dictionary as a method. Thus, a language dictionary suitable for a user of the speech recognition device can be prepared.
COPYRIGHT: (C)2005,JPO&NCIPI
摘要翻译：要解决的问题：为了提供能够通过有效地使用存储器而能够灵活地与应用兼容的语音识别装置，因为尽管传统语音识别装置已经利用网络语言字典和统计语言字典，它们既有优缺点，也没有实现实用性能，而且尽管还提供了组合它们的方法，但还没有获得足够的结果。解决方案：语音识别装置一起使用网络语法语言字典和统计语言字典，并提取具有多个频率信息的词汇，以从后者语言字典连接到包括在前语言词典中的词汇A，以及将其添加到前面的语言字典作为一种方法。因此，可以准备适合于语音识别装置的用户的语言字典。版权所有（C）2005，JPO＆NCIPI

44. 发明专利

JP2005173134A Speech recognition apparatus 审中-公开
标题翻译：语音识别装置
公开(公告)号：JP2005173134A
公开(公告)日：2005-06-30
申请号：JP2003412250
申请日：2003-12-10
申请人： Nissan Motor Co Ltd , 日産自動車株式会社
发明人： KAMINUMA MITSUNOBU , CHINOMI SATOSHI
IPC分类号： G10L15/18 , G10L15/00 , G10L15/08 , G10L15/10 , G10L15/28
摘要： PROBLEM TO BE SOLVED: To provide a speech recognition apparatus which creates an appropriate language dictionary by dynamically changing the language dictionary using information acquired by a communication means.
SOLUTION: The speech recognition apparatus has a speech input means for inputting speech, a storage means for storing the language dictionary which is compared with the speech and evaluated, a speech recognition means for evaluating the speech by the language dictionary, the communication means, a next utterance prediction means for predicting the next utterance from information acquired by the communication means and a changing means for changing the contents of the language dictionary based on the next utterance predicted by the next utterance prediction means.
COPYRIGHT: (C)2005,JPO&NCIPI
摘要翻译：要解决的问题：提供一种语音识别装置，其通过使用由通信装置获取的信息动态地改变语言词典来创建适当的语言词典。解决方案：语音识别装置具有用于输入语音的语音输入装置，用于存储与语音进行比较并评估的语言字典的存储装置，用于通过语言字典评估语音的语音识别装置，通信意味着下一个话音预测装置，用于根据通信装置获取的信息预测下一个话语;以及改变装置，用于根据下一个话音预测装置预测的下一个发音改变语言词典的内容。版权所有（C）2005，JPO＆NCIPI

45. 发明专利

JP2005037803A Speech recognition device 有权
标题翻译：语音识别装置
公开(公告)号：JP2005037803A
公开(公告)日：2005-02-10
申请号：JP2003276529
申请日：2003-07-18
申请人： Nissan Motor Co Ltd , 日産自動車株式会社
发明人： KAMINUMA MITSUNOBU , NAKAYAMA OKIHIKO
IPC分类号： G01C21/00 , G10L15/00 , G10L15/06 , G10L15/10 , G10L15/28
摘要： PROBLEM TO BE SOLVED: To provide a speech recognition device to which an updating method for a language dictionary with which a vocabulary before replacement can be recognized even after the replacement by the update of the language dictionary and memory capacity, recognition rate or recognition time are not affected, can be applied. SOLUTION: The speech recognition device is principally constituted by installing a first memory means to use the language dictionary for recognition processing and a second memory means to use the language dictionary for the update processing, and storing, when the contents of the vocabularies stored in both the memory means are the same object, both the vocabularies in the first memory means together with an operation command indicating that the contents of of the vocabularies are the same object. Also, the constitution to control the updating method according to the use history of the vocabulary and the lapse of time after the update is employed for the dictionary updating. Consequently, the problem that when the language dictionary is updated in the speech recognition device, a user is requested to store the new information each time the update is performed and as a solution for the same, a method of previously storing both of the new and the old information is proposed but in such a case, the capacity of the storage device is increased and this causes the worry about the influence on a conversion speed and conversion accuracy, can be solved by the speech recognition device of this invention. Further, the speech recognition device includes the updating method for the language dictionary, with which a load on the user is reduced. COPYRIGHT: (C)2005,JPO&NCIPI
摘要翻译：要解决的问题：提供一种语音识别装置，即使在通过更新语言字典和存储器容量的替换之后也可以识别用于替换之前的词汇的语言字典的更新方法，识别率或识别时间不受影响，可以应用。解决方案：语音识别装置主要通过安装第一存储装置来使用用于识别处理的语言字典和第二存储装置来使用用于更新处理的语言字典，并且当词汇的内容存储在两个存储器装置中的是相同的对象，第一存储器中的词汇表都与指示词表的内容是相同对象的操作命令一起。此外，采用根据词汇的使用历史和更新之后的时间流逝来控制更新方法的结构用于字典更新。因此，当在语音识别装置中更新语言字典时，要求用户每次执行更新时存储新信息，并且作为其解决方案的问题，一种先前存储新的和提出了旧的信息，但是在这种情况下，存储装置的容量增加，这导致担心对转换速度和转换精度的影响，可以通过本发明的语音识别装置来解决。此外，语音识别装置包括语言字典的更新方法，用户的负载减少。版权所有（C）2005，JPO＆NCIPI

46. 发明专利

JP2004333870A Speech recognition device for traveling object 有权
标题翻译：用于旅行对象的语音识别装置
公开(公告)号：JP2004333870A
公开(公告)日：2004-11-25
申请号：JP2003129740
申请日：2003-05-08
申请人： Nissan Motor Co Ltd , 日産自動車株式会社
发明人： KAMINUMA MITSUNOBU , RI AKINOBU
IPC分类号： G10L15/00 , G10L15/06 , G10L15/18 , G10L15/22 , G10L15/28
CPC分类号： G10L15/197 , G10L15/183
摘要： PROBLEM TO BE SOLVED: To realize a recognition device that can have a high degree of freedom of grammatical input to a statistical language dictionary and dynamically realize high recognition performance like a network language dictionary while both the dictionary are mainly proposed for speech recognition.
SOLUTION: An object to be applied is considered to be for a traveling object and simple speaking prediction is previously performed according to network grammar. A method of dynamically correcting the probability of concatenated words in the sequential and statistical language dictionary is based upon. A plurality of small-scale network grammar language dictionaries are installed as an embodied device and switched by contents, and necessary nodes of a large-scale network grammar language dictionary are selected. Here, the network language dictionary has limitations on grammar which can be inputted, but excellent recognition precision can be realized. The statistical language dictionary, on the other hand, has less grammatical restrictions, but speech recognition precision is low. In this embodiment, such problems are solved.
COPYRIGHT: (C)2005,JPO&NCIPI
摘要翻译：要解决的问题：为了实现能够对统计语言字典具有高自由度的语法输入的识别装置，并且动态地实现诸如网络语言字典之类的高识别性能，而这两种字典主要用于语音识别。解决方案：要应用的对象被认为是用于旅行对象，并且根据网络语法预先执行简单的说话预测。在顺序和统计语言字典中动态地校正连接词的概率的方法是基于的。安装多个小规模网络语法语言字典作为具体的设备，并通过内容进行切换，并选择大规模网络语法语言字典的必要节点。这里，网络语言字典对可以输入的语法有限制，但是可以实现良好的识别精度。另一方面，统计语言字典语法限制较少，但语音识别精度较低。在本实施例中，解决了这些问题。版权所有（C）2005，JPO＆NCIPI

47. 发明专利

JP2004251769A Information providing device 审中-公开
公开(公告)号：JP2004251769A
公开(公告)日：2004-09-09
申请号：JP2003042711
申请日：2003-02-20
申请人： Nissan Motor Co Ltd , 日産自動車株式会社
发明人： KAMINUMA MITSUNOBU , YAMAZAKI MASARU , ONO TAKESHI
IPC分类号： B60R21/00 , B60R21/34 , G01S7/04 , G01S13/93 , G08B21/00
摘要： PROBLEM TO BE SOLVED: To provide information with sound regarding a relative positional relation between own vehicle (moving body) and an obstacle. SOLUTION: This device comprises a relative position detection means 1 detecting the relative position of the obstacle to its own vehicle (moving body), and a sounding means 2 generating sound at least two sounding source, wherein the sounding means 2 determines one or two or more positions corresponding to its own vehicle and comprises a 1st sounding part 21 sounding 1st sound sources arranged at the determined points, and determines one or two or more positions corresponding to the obstacles, and comprises a 2nd sounding part 22 sounding 2nd sound sources arranged at the determined points. COPYRIGHT: (C)2004,JPO&NCIPI

48. 发明专利

JP2003035550A Navigation system 有权
标题翻译：导航系统
公开(公告)号：JP2003035550A
公开(公告)日：2003-02-07
申请号：JP2001223047
申请日：2001-07-24
申请人： Nissan Motor Co Ltd , 日産自動車株式会社
发明人： KAMINUMA MITSUNOBU
IPC分类号： G01C21/00 , G08G1/0969 , H04S1/00 , H04S7/00
摘要： PROBLEM TO BE SOLVED: To reduce troublesome feeling of a driver when a virtual sound source is presented audibly, and to accurately present the position of a guidance route to the driver.
SOLUTION: The system is provided with sound output means 4, 8 and 7 which output sound, a current position detecting section 1 which detects current position information representing the current position of an own vehicle, an input section 2 which inputs destination information representing the destination of the own vehicle, a course determining/detecting section 11 which searches the guidance route through which the own vehicle reaches the destination being input in the input section 2 from the current position being input in the input section 2, a target spot determining section 12 which determines a spot to be targeted for the driver as a target spot in the guidance route searched by the course determining/detecting section 11, and a virtual sound source signal outputting/judging section 13 and a virtual sound source signal generating section 14 which calculate a relative position between the target spot being determined in the target spot determining section 12 and the current position being input in the input section 2, fix and set at least one virtual sound source at the target spot being determined in the target spot determining section 12, and generate/output a virtual sound source signal presenting the guidance route to the driver from the virtual sound source on the basis of the relative position.
COPYRIGHT: (C)2003,JPO
摘要翻译：要解决的问题：为了减少虚拟声源可听见地呈现驾驶员的烦琐感，并准确地向驾驶者呈现引导路线的位置。解决方案：系统设有输出声音的声音输出装置4,8和7，检测表示本车当前位置的当前位置信息的当前位置检测部分1，输入表示目的地的目的地信息的输入部分2 ，从输入部2中输入的当前位置搜索本车到达目的地的引导路线的路线确定/检测部11，目标点确定部12 其将由驾驶员定位的点作为由路线确定/检测部分11搜索的引导路线中的目标点，虚拟声源信号输出/判断部分13和虚拟声源信号产生部分14计算在目标点确定部分12中确定的目标点与当前位置之间的相对位置在输入部分2中输入，在目标点确定部分12确定的目标点处固定和设置至少一个虚拟声源，并且从目标点确定部分12中向驾驶员生成/输出呈现引导路线的虚拟声源信号虚拟声源在相对位置的基础上。

49. 发明专利

JP2014046820A Driver's property estimation system 审中-公开
标题翻译：驾驶员财产估算系统
公开(公告)号：JP2014046820A
公开(公告)日：2014-03-17
申请号：JP2012191398
申请日：2012-08-31
申请人： Nissan Motor Co Ltd , 日産自動車株式会社
发明人： NAGAI YUKIKO , KAMINUMA MITSUNOBU , BUE STEFAN , TANIDA SHIGEHITO
IPC分类号： B60W40/09 , G08G1/16
摘要： PROBLEM TO BE SOLVED: To provide a driver's property estimation system capable of estimating a fundamental property of a driver.SOLUTION: A scene where a vehicle is rolling is estimated based on information representing an ambient situation imaged by a camera 10, and a driver's driving action is detected based on vehicle information detected by a vehicle signal detection unit 11. Based on the driving action which the driver has performed in the estimated scene, the driver's property is estimated through correlation with a property tendency label stored in a driver's property database 20. Eventually, the driver's property can be properly estimated.
摘要翻译：要解决的问题：提供能够估计驾驶员的基本属性的驾驶员属性估计系统。解决方案：基于由照相机10成像的环境情况的信息和驾驶员驾驶来估计车辆滚动的场景基于车辆信号检测单元11检测到的车辆信息检测动作。根据驾驶员在估计场景中进行的驾驶动作，通过与存储在驾驶员属性数据库20中的属性倾向标签相关联来推定驾驶员的属性最终，司机的财产可以正确估计。

50. 发明专利

JP2012138092A Acoustic sense monitoring device for vehicle 审中-公开
公开(公告)号：JP2012138092A
公开(公告)日：2012-07-19
申请号：JP2012010065
申请日：2012-01-20
申请人： Nissan Motor Co Ltd , 日産自動車株式会社
发明人： YAMAZAKI MASARU , KAMINUMA MITSUNOBU
IPC分类号： G08G1/16 , G01C21/26 , H04R3/00 , H04R5/027 , H04S5/02 , H04S7/00
摘要： PROBLEM TO BE SOLVED: To allow a vehicle occupant to hear external sound detected by adapting to a traveling state of one's own vehicle.SOLUTION: Reproduced sound based on collected sound around a vehicle is generated. A traveling state of one's own vehicle is determined based on detected information related to the one's own vehicle. The reproduced sound output based on the determined traveling state of the one's own vehicle is controlled.

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