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. 发明专利

JP2005283531A Vehicle-mounted device and data-preparing device 审中-公开
标题翻译：车辆安装设备和数据准备设备
公开(公告)号：JP2005283531A
公开(公告)日：2005-10-13
申请号：JP2004101851
申请日：2004-03-31
申请人： Equos Research Co Ltd , 株式会社エクォス・リサーチ
发明人： HIRANO TADASHI , KUBOTA TOMOKI , MORITA HIDEAKI , NOMOTO YOICHI , OKABE HIDEFUMI , ADACHI KAZUHIDE
IPC分类号： G01C21/00 , G08G1/0969
摘要： PROBLEM TO BE SOLVED: To reduce the sense of incongruity for a passenger, with respect to a vehicle to outputted voice content. SOLUTION: When a scenario preparer can assume the person who sits on each seat, speech (polite and frank expression) differing for each seat is set. When a person who sits on the seat cannot be foreseen, a scenario editor sets the type of a passenger (age, sex, social status), and speech corresponding to each type is created for each scene. An agent device specifies the type and seat position of a passenger, in advance, for example by asking questions and obtaining responses to the questions. A scenario for satisfying the automatic start conditions of the scenario is executed. When there are a plurality of speeches outputted at each scene and the seat is specified for each speech, speech is outputted from a loudspeaker at a position, corresponding to the specified seat. When the type of passengers to each speech is set, speech is outputted from a loudspeaker, corresponding to the type and seat position of the passenger that are specified, in advance. COPYRIGHT: (C)2006,JPO&NCIPI
摘要翻译：要解决的问题：为了减少乘客对于车辆输出语音内容的不协调感。
解决方案：当场景准备员可以假设坐在每个座位上的人时，设置每个座位不同的言语（礼貌和坦率的表达）。当坐在座位上的人不能预见时，场景编辑者设定乘客的类型（年龄，性别，社会地位），并为每个场景创建与每种类型相对应的语音。代理设备提前指定乘客的类型和座位位置，例如通过提问和获得对问题的回答。执行用于满足场景的自动启动条件的场景。当在每个场景中输出多个演讲并且为每个语音指定座席时，在与指定的座位对应的位置处，从扬声器输出语音。当设定每个语音的乘客的类型时，预先从指定的乘客的类型和座位位置从扬声器输出语音。版权所有（C）2006，JPO＆NCIPI

52. 发明专利

JP2005215411A Voice synthesizing apparatus and method 审中-公开
标题翻译：语音合成设备和方法
公开(公告)号：JP2005215411A
公开(公告)日：2005-08-11
申请号：JP2004022977
申请日：2004-01-30
申请人： Equos Research Co Ltd , 株式会社エクォス・リサーチ
发明人： NOMOTO YOICHI , MORITA HIDEAKI , KUBOTA TOMOKI , OKABE HIDEFUMI
IPC分类号： G10L13/08 , G10L13/00 , G10L13/06
摘要： PROBLEM TO BE SOLVED: To provide a voice synthesizing apparatus and a voice synthesizing method in which voice having good feeling for listening can be synthesized without requiring an enormous amount of voice data. SOLUTION: In the voice synthesizing apparatus, input text data are analyzed to detect a postpositional word (S101), postpositional word voice data of the postpositional word corresponding to the mora number and the accent type of a word to which the postpositional word detected in the text data is placed, are determined (S105), postpositional word voice data of the determined postpositional word are selected from a voice database (S107), and voice data corresponding to the text data that include the postpositional word voice data of the selected postpositional word are constituted and output from a voice output device (S109). Thus, postpositional word voice data of the postpositional word corresponding to the mora number and the accent type of the word which is placed immediately before the postpositional word, are determined and the data are selected from "the voice database which stores the postpositional word voice data of the postpositional word classified by the mora number and the accent type of the word that is placed immediately before the postpositional word". COPYRIGHT: (C)2005,JPO&NCIPI
摘要翻译：要解决的问题：提供一种语音合成装置和语音合成方法，其中可以合成具有良好听力感的语音，而不需要大量语音数据。解决方案：在语音合成装置中，分析输入文本数据以检测后置词（S101），对应于后缀词的后缀词的后置词语音数据和词后缀词的重音类型（S105），从语音数据库中选择所确定的后置词的后置词语音数据（S107），对应于包括文本数据的文本数据的文本数据的语音数据从语音输出装置构成并输出所选择的后置字（S109）。因此，确定与紧随在后置字之前放置的字的mora数和重音类型相对应的后置词的后置词语音数据，并且从“存储后置词语音数据的语音数据库”中选择数据的后缀词被分类为词义数字和立即放置在词后面的单词的重音类型“。版权所有（C）2005，JPO＆NCIPI

53. 发明专利

JP2005215287A Voice synthesizing apparatus and method 有权
标题翻译：语音合成设备和方法
公开(公告)号：JP2005215287A
公开(公告)日：2005-08-11
申请号：JP2004021403
申请日：2004-01-29
申请人： Equos Research Co Ltd , 株式会社エクォス・リサーチ
发明人： NOMOTO YOICHI , KUBOTA TOMOKI , MORITA HIDEAKI , OKABE HIDEFUMI
IPC分类号： G10L13/08 , G10L13/06
摘要： PROBLEM TO BE SOLVED: To provide a voice synthesizing apparatus and a voice synthesizing method with which voice having good feeling for listening can be synthesized without requiring an enormous amount of data.
SOLUTION: In the voice synthesizing apparatus, input text data are divided into prescribed clauses by a text analysis process S101, accent types of the prescribed clauses are analyzed by an accent analysis program process S103 and comparison is made to determine whether the accent types of voice data corresponding to the word in the prescribed clause and the accent types of the word portion that constitutes the accent types of the prescribed clauses are coincident with each other or not by an accent type comparison process S107. When the accent types of the both are not coincident with each other (i.e., the case of "No" in S107), accent conversion is conducted for the voice data by a phoneme data substitution process S113 so that it is made to coincide with the analyzed accent type, and voice data corresponding to the text data that include the accent-converted voice data are constituted and output from a voice output device by a voice data output process S117.
COPYRIGHT: (C)2005,JPO&NCIPI
摘要翻译：要解决的问题：提供一种语音合成装置和语音合成方法，可以在不需要大量数据的情况下合成具有良好听力感的语音。解决方案：在语音合成装置中，通过文本分析处理S101将输入文本数据分成规定的子句，通过重音分析程序处理S103分析规定子句的重音类型，并进行比较以确定是否重音与规定条款中的单词相对应的语音数据的类型和构成规定子句的重音类型的单词部分的重音类型彼此一致或不通过重音类型比较处理S107。当两者的重音类型彼此不一致时（即，S107中为“否”的情况）），通过音素数据替换处理S113对语音数据进行重音转换，使得它与分析的重音类型，并且通过语音数据输出处理S117从语音输出装置构成并输出与包含重音转换的语音数据的文本数据相对应的语音数据。版权所有（C）2005，JPO＆NCIPI

54. 发明专利

JP2004025942A Parking operation support device 审中-公开
公开(公告)号：JP2004025942A
公开(公告)日：2004-01-29
申请号：JP2002182097
申请日：2002-06-21
申请人： Equos Research Co Ltd , 株式会社エクォス・リサーチ
发明人： KUBOTA TOMOKI , MORITA HIDEAKI , OKABE HIDEFUMI
IPC分类号： B60R21/00 , B62D15/02
CPC分类号： B62D15/028
摘要： PROBLEM TO BE SOLVED: To provide a parking operation support device for relieving a burden of parking operation. SOLUTION: A parking area 33 and an obstacle 32 are detected by detecting right and left distances between the obstacle and a vehicle using right and left ultrasonic sensors 21 positioned on the front side of a vehicle, and the distances are displayed on a parking operation support screen of a display unit along with the position of an own vehicle. To move the vehicle to the parking area, an optimum vehicle position to be moved from a present vehicle position is calculated and its image is displayed as a guidance area 34. A driver can grasp an appropriate position to be moved next by displaying an image of the guidance area along with the position of the own vehicle, the parking area, and the obstacle. Additionally, a moving locus is displayed as a guideline when the vehicle moves forward/backward at a present steering angle. The driver can obtain an optimum steering angle by steering operation so that the guideline may meet the guidance area. A vehicle symbol 31 is always displayed in a heading-up state. COPYRIGHT: (C)2004,JPO

55. 发明专利

JP2005215288A Voice synthesizing apparatus and method 审中-公开
标题翻译：语音合成设备和方法
公开(公告)号：JP2005215288A
公开(公告)日：2005-08-11
申请号：JP2004021404
申请日：2004-01-29
申请人： Equos Research Co Ltd , 株式会社エクォス・リサーチ
发明人： NOMOTO YOICHI , KUBOTA TOMOKI , MORITA HIDEAKI , OKABE HIDEFUMI
IPC分类号： G10L13/08 , G10L13/00 , G10L13/06
摘要： PROBLEM TO BE SOLVED: To provide a voice synthesizing apparatus and a voice synthesizing method with which voice having good feeling for listening is synthesized without requiring an enormous amount of voice data.
SOLUTION: In the voice synthesizing apparatus, input text data are analyzed on the basis of the word data of a text analysis dictionary (S101) and an unknown word that is not stored in the text analysis dictionary is detected. Then, read data to the unknown word are obtained (S105), a mora number and an accent type of the unknown word are determined on the basis of the obtained read data of the unknown word (S107), voice data of the unknown word are generated on the basis of the determined mora number and the accent type (S115), and voice data corresponding to the text data including the voice data of the generated unknown word are constituted and the data are output from a voice outputting device (S117).
COPYRIGHT: (C)2005,JPO&NCIPI
摘要翻译：要解决的问题：提供一种语音合成装置和语音合成方法，在不需要大量语音数据的情况下合成具有良好听觉感的语音。解决方案：在语音合成装置中，基于文本分析词典（S101）的字数据分析输入文本数据，并检测未存储在文本分析词典中的未知字。然后，获取对未知字的读取数据（S105），根据获得的未知字的读取数据（S107）确定未知字的mora数和重音类型，未知字的语音数据为（S115），并且构成对应于包括所生成的未知字的语音数据的文本数据的语音数据，并从声音输出装置输出数据（S117）。版权所有（C）2005，JPO＆NCIPI

56. 发明专利

JP2005062363A Audio data output device 有权
标题翻译：音频数据输出设备
公开(公告)号：JP2005062363A
公开(公告)日：2005-03-10
申请号：JP2003290848
申请日：2003-08-08
申请人： Equos Research Co Ltd , 株式会社エクォス・リサーチ
发明人： KUBOTA TOMOKI , OKABE HIDEFUMI
IPC分类号： G10L13/04
摘要： PROBLEM TO BE SOLVED: To improve an audio data retrieval speed by decreasing a frequency of data access in retrieval.
SOLUTION: Cell data of prescribed size are generated by adding a margin area to audio waveform data and cell data of the same size are gathered and stored, group by group. A cell group code and the order of cell storage in a cell group are regarded as the ID number of audio data. Further, an index table consisting of cell group codes and head addresses is stored in a RAM (1 to 4) functioning as a cache memory. Then when audio waveform data corresponding to an ID number A2m of a voice are retrieved, a cell group code A2 is obtained from the ID number first and its head address a2 is retrieved from the index table. Then a cell array number (m) and the cell group code A2 are used to calculate the address of (m)th audio data from the head address a2. Audio waveform data are read out of calculated addresses and combined in the order of an ID number array to be outputted.
COPYRIGHT: (C)2005,JPO&NCIPI
摘要翻译：要解决的问题：通过减少检索中的数据访问频率来提高音频数据检索速度。解决方案：通过向音频波形数据添加边缘区域，并逐组收集和存储相同大小的单元格数据，生成规定大小的单元格数据。小区组代码和小区组中的小区存储顺序被认为是音频数据的ID号。此外，由小区组代码和头地址组成的索引表存储在用作高速缓冲存储器的RAM（1至4）中。然后，当检索到对应于语音的ID号码A2m的音频波形数据时，从ID号首先获得小区组代码A2，并从索引表中检索其起始地址a2。然后使用单元阵列号（m）和单元组代码A2来从头地址a2计算第（m）个音频数据的地址。从计算出的地址中读出音频波形数据，并以要输出的ID号码阵列的顺序进行组合。版权所有（C）2005，JPO＆NCIPI

57. 发明专利

JP2008109705A Information communication equipment for traveling body 有权
标题翻译：信息通信设备的行车
公开(公告)号：JP2008109705A
公开(公告)日：2008-05-08
申请号：JP2007333357
申请日：2007-12-25
申请人： Equos Research Co Ltd , 株式会社エクォス・リサーチ
发明人： KUBOTA TOMOKI , MORITA HIDEAKI , OKABE HIDEFUMI
IPC分类号： G08G1/09 , H04B7/26 , H04W72/04
摘要： PROBLEM TO BE SOLVED: To transmit and receive information by preventing interference without causing increase in facility load, even when a plurality of transmission spots of vehicles exit closely. SOLUTION: A transmission vehicle A transmits information with a frequency f1 set by a grid G1. A transmission vehicle B transmits information with a second frequency f2 set by a grid G2. A reception vehicle M receives information with the frequency set, by the grid selected by a user from among the grids G1 and G2. For example, when a transmission frequency is set to the frequency of the grid G2, the reception vehicle M receives the information transmitted from the transmission vehicle A. The channel for transmission and reception is set for every grid; and even when information is transmitted from the plurality of transmission and reception vehicles, interference does not occur. COPYRIGHT: (C)2008,JPO&INPIT
摘要翻译：要解决的问题：即使当车辆的多个传播点紧密地离开时，通过防止干扰而不引起设备负载的增加来发送和接收信息。解决方案：传输车辆A发送具有由网格G1设置的频率f1的信息。传输车辆B发送具有由网格G2设置的第二频率f2的信息。接收车辆M通过由用户从网格G1和G2中选择的网格来接收频率设定的信息。例如，当将发送频率设定为电网G2的频率时，接收车辆M接收从发送车辆A发送的信息。为每个电网设定发送和接收信道; 并且即使当从多个发送和接收车辆发送信息时，也不会发生干扰。版权所有（C）2008，JPO＆INPIT

58. 发明专利

JP2006164197A Running support system 审中-公开
标题翻译：运行支持系统
公开(公告)号：JP2006164197A
公开(公告)日：2006-06-22
申请号：JP2004358878
申请日：2004-12-10
申请人： Equos Research Co Ltd , 株式会社エクォス・リサーチ
发明人： HORI KOJI , OKABE HIDEFUMI , MASUDA YOSHITAKA
IPC分类号： G08G1/16 , B60R21/00
摘要： PROBLEM TO BE SOLVED: To support a driver in running by notifying, at merging from a first road to a second road, a running vehicle on the second road, without requiring a special infrastructure.
SOLUTION: Accurate positions and shapes of zebra zones positioned at junctions such as ramp ways are stored in a high resolution road map DB as object information. When an associated vehicle is in any junction, the accurate position and orientation of the vehicle are detected from a zebra zone recognized from an image picked up by a front imaging device and the object information. According to the accurate position and orientation of the vehicle and road information such as the merging angle, curve curvature and number of lanes of the merging point, an area to be recognized (for example, including a merging lane and excluding a passing lane) is specified from a side back image picked by a side back imaging device. The position and size of each side back vehicle running on the merging lane are recognized from the image in the specified area and notified to the driver.
COPYRIGHT: (C)2006,JPO&NCIPI
摘要翻译：要解决的问题：在不需要特殊基础设施的情况下，通过在第一道路到第二道路合并第二道路上的行驶车辆来通知通知的驾驶员。
解决方案：位于交叉点（如斜坡路）上的斑马带的精确位置和形状作为对象信息存储在高分辨率路线图DB中。当相关联的车辆处于任何交界处时，从由从前成像设备拾取的图像和对象信息中识别的斑马瓦区域检测车辆的准确位置和方位。根据车辆的准确位置和方向以及合并角度，曲线曲率和合并路线数等道路信息，要识别的区域（例如包括合并车道，不包括通行车道）为由侧背成像装置拾取的侧背图像指定。从指定区域的图像中识别在合并车道上行驶的每个侧向后车辆的位置和大小，并通知驾驶员。版权所有（C）2006，JPO＆NCIPI

59. 发明专利

JP2004287192A Device and program for editing synthesized speech 审中-公开
公开(公告)号：JP2004287192A
公开(公告)日：2004-10-14
申请号：JP2003080665
申请日：2003-03-24
申请人： Equos Research Co Ltd , 株式会社エクォス・リサーチ
发明人： KUBOTA TOMOKI , MORITA HIDEAKI , NOMOTO YOICHI , OKABE HIDEFUMI , FUJII KAZUAKI
IPC分类号： G10L13/00 , G10L13/04 , G10L21/06
摘要： PROBLEM TO BE SOLVED: To simply display a display regarding editing of a synthesized speech. SOLUTION: A voiced part of the synthesized speech is displayed with a framed word and a silent part is displayed with a segment (silent segment). In the frame of the voiced part, the reading of the word corresponding to the voiced part is displayed as a text. Consequently, a user is able to easily understand what unit of unit speech data synthesized speech data and synthesized speech data to be edited are composed. A line whose length corresponds to the length of the silent part is displayed between frames, so the user is able to sensually grasp whether there is a silent part and its length. Words of unit speech data are categorized (for example, by parts of speech) and lengths of following silent parts are automatically set by categories. The lengths of the silent parts can be edited by variation, deletion, etc. COPYRIGHT: (C)2005,JPO&NCIPI

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