专利快速检索-快速检索全球专利，免费商用专利数据库-IPRDB

11. 发明专利

JP2008026460A Encoding device, method therefor, program, and recording medium 有权
标题翻译：编码设备，其方法，程序和记录介质
公开(公告)号：JP2008026460A
公开(公告)日：2008-02-07
申请号：JP2006196810
申请日：2006-07-19
申请人： Nippon Telegr & Teleph Corp , 日本電信電話株式会社
发明人： HARADA NOBORU , KAMAMOTO MASARU , MORIYA TAKEHIRO
IPC分类号： G10L19/22 , H03M7/36 , H04N19/102 , H04N19/149 , H04N19/156 , H04N19/172 , H04N19/196 , H04N19/50
摘要： PROBLEM TO BE SOLVED: To perform an encoding capable of sufficiently making use of the arithmetic processing capability of a decoding device without impairing real-time characteristics of decoding processing. SOLUTION: A tolerance of a parameter specifying encoding processing of a current frame is extracted so that a maximum arithmetic processing amount within the tolerance has a value not more than a maximum value of an arithmetic processing amount of decoding processing of the current frame of a decoding device specified with the difference between a first index value corresponding to a decoding operation amount by which the decoding device can perform processing in a second section comprising a current frame to be encoded and a first section which is the time section right before it and a second index value corresponding to an arithmetic processing amount the decoding device needs to have to decode a code obtained by actually encoding an input signal in the first section. COPYRIGHT: (C)2008,JPO&INPIT
摘要翻译：要解决的问题：执行能够充分利用解码装置的算术处理能力的编码，而不损害解码处理的实时特性。解决方案：提取指定当前帧的编码处理的参数的公差，使得容差内的最大算术处理量具有不大于当前帧的解码处理的算术处理量的最大值的值解码装置指定的解码装置，其具有与解码装置可以在包括当前要编码的帧的第二部分中执行处理的解码操作量相对应的第一索引值与作为要编码的当前帧之间的时间段的第一部分之间的差以及对应于解码装置需要对通过在第一部分中实际编码输入信号而获得的代码进行解码的算术处理量的第二索引值。版权所有（C）2008，JPO＆INPIT

12. 发明专利

JP2007318691A Apparatus and method for determining linear prediction model degree, program thereof and recording medium 有权
标题翻译：用于确定线性预测模型程度，程序和记录介质的装置和方法
公开(公告)号：JP2007318691A
公开(公告)日：2007-12-06
申请号：JP2006148919
申请日：2006-05-29
申请人： Nippon Telegr & Teleph Corp , 日本電信電話株式会社
发明人： MORIYA TAKEHIRO , HARADA NOBORU , KAMAMOTO MASARU
IPC分类号： H03M7/30 , G10L11/00 , G10L19/06
摘要： PROBLEM TO BE SOLVED: To determine an optimal model degree of a linear prediction model without actually determining a code amount for each model degree. SOLUTION: A PARCOR coefficient for each linear prediction model to a model degree (m) and a forward prediction error sequence and/or a backward prediction error sequence in the model degree (m) are calculated, respectively from an input signal. A code amount of a prediction error waveform that is an error between a predictive value of the linear prediction model of the model degree (m) and the input signal, is then estimated from the forward prediction error sequence and/or the backward prediction error sequence in the model degree (m). An individual code amount corresponding to respective PARCOR coefficients to the model degree (m) or an entire code amount corresponding to all the PARCOR coefficients is calculated. Regarding a plurality of model degrees, a total code amount of a total of entire code amounts in the respective model degrees (m) or of individual code amounts to the model degree (m) and estimated code amounts of the prediction error waveform, is found and one model degree is determined from the total code amount. COPYRIGHT: (C)2008,JPO&INPIT
摘要翻译：要解决的问题：确定线性预测模型的最佳模型度，而不确定每个模型度的代码量。解决方案：分别从输入信号计算模型度（m）中的每个线性预测模型对模型度（m）和前向预测误差序列和/或后向预测误差序列的PARCOR系数。然后，从正向预测误差序列和/或反向预测误差序列估计作为模型度（m）的线性预测模型的预测值与输入信号之间的误差的预测误差波形的编码量在模型度（m）。计算对应于与模型度（m）的相应PARCOR系数相对应的单个代码量或对应于所有PARCOR系数的整个代码量。关于多个模型度，找到相应模型度（m）或各个代码中的总代码量的总代码量相当于模型度（m）和预测误差波形的估计代码量并从总代码量确定一个模型度。版权所有（C）2008，JPO＆INPIT

13. 发明专利

JP2007286146A Adaptive block length encoding device, its method, program and recording medium 有权
标题翻译：自适应长度编码设备，其方法，程序和记录介质
公开(公告)号：JP2007286146A
公开(公告)日：2007-11-01
申请号：JP2006110511
申请日：2006-04-13
申请人： Nippon Telegr & Teleph Corp , 日本電信電話株式会社
发明人： MORIYA TAKEHIRO , HARADA NOBORU , KAMAMOTO MASARU
IPC分类号： G10L19/14 , H03M7/36
摘要： PROBLEM TO BE SOLVED: To reduce an arithmetic amount for encoding, in an adaptive block length encoding in which a frame is encoded by hierarchically dividing it into a plurality of blocks. SOLUTION: A residual code amount estimation value is calculated for each block which is obtained by hierarchically dividing the frame into the plurality of blocks. Then, when a period in the frame is composed of different blocks in two layers or more, a size of a sum of the residual code amount estimation value in each layer corresponding to the period is compared. By using this comparison result, a set of blocks for composing the frame in which the sum of the code amount corresponding to each block is the smallest, is selected, and a prediction residual signal in which a residual encoding section corresponds to each selected block, is encoded. COPYRIGHT: (C)2008,JPO&INPIT
摘要翻译：要解决的问题：为了减少编码的算术量，在自适应块长度编码中，其中帧被分级地分成多个块来编码。解决方案：针对通过将帧分层划分成多个块而获得的每个块来计算残差代码量估计值。然后，当帧中的周期由两层或更多层的不同块组成时，比较与该周期对应的每层中的残留码量估计值之和的大小。通过使用该比较结果，选择用于构成其中与每个块相对应的代码量的和最小的帧的一组块，以及残差编码部分对应于每个选择的块的预测残差信号，被编码。版权所有（C）2008，JPO＆INPIT

14. 发明专利

JP2007150744A Signal multiplexing device, device for separating multiple signal and their method 有权
标题翻译：信号多路复用器件，用于分离多个信号的器件及其方法
公开(公告)号：JP2007150744A
公开(公告)日：2007-06-14
申请号：JP2005342579
申请日：2005-11-28
申请人： Nippon Telegr & Teleph Corp , 日本電信電話株式会社
发明人： MORIYA TAKEHIRO , HARADA NOBORU , KAMAMOTO MASARU , MATSUTANI YASUYUKI
IPC分类号： H04J3/00
摘要： PROBLEM TO BE SOLVED: To provide a multiplexing device having a simple constitution, minimizing the number of the signal lines of a transmission system, and being capable of transmitting a signal at a high speed, by making the most of the feature of an over-sampling low-amplitude encoding technique and a device for separating a multiple signal.
SOLUTION: In the signal multiplexing device, a large number of elements 10a changing an observation signal (such as an output signal from a microphone) into a digital signal by the over-sampling low-amplitude encoding technique are dispersed and arranged and a common signal line 11 connected to these dispersed elements is fitted. In the signal multiplexing device, an element control unit 70a is mounted being fitted at every large number of the elements and outputting a digital output signal to the common signal line 11 at every one period of an over-sampling. The device for separating the multiple signal has an element control unit 120a generating holding signals fitted at every element holding the multiplexed signal at every one period of the over-sampling on the common signal line 11, and a plurality of the elements (such as a loudspeaker) decoding the held digital signals to analog signals in the signal multiplexing device.
COPYRIGHT: (C)2007,JPO&INPIT
摘要翻译：要解决的问题：为了提供具有简单结构的多路复用装置，使传输系统的信号线的数量最小化，并且能够高速地发送信号，通过充分利用过采样低幅度编码技术和用于分离多信号的装置。解决方案：在信号复用装置中，通过过采样低幅度编码技术将大量将观测信号（例如麦克风的输出信号）改变为数字信号的元件10a被分散和布置，连接到这些分散元件的公共信号线11被装配。在信号多路复用装置中，元件控制单元70a安装在每个大量的元件上，并且在过采样的每一个周期将数字输出信号输出到公共信号线11。用于分离多信号的装置具有一个元件控制单元120a，产生在公共信号线11上的过采样的每一个周期处保持多路复用信号的每个元件处装配的保持信号，以及多个元件扬声器）将保持的数字信号解码为信号多路复用装置中的模拟信号。版权所有（C）2007，JPO＆INPIT

15. 发明专利

JP2007150743A Transmitter 有权
标题翻译：发射机
公开(公告)号：JP2007150743A
公开(公告)日：2007-06-14
申请号：JP2005342578
申请日：2005-11-28
申请人： Nippon Telegr & Teleph Corp , 日本電信電話株式会社
发明人： MORIYA TAKEHIRO , HARADA NOBORU , KAMAMOTO MASARU
IPC分类号： H04R3/00 , H04B7/26 , H04R1/40
摘要： PROBLEM TO BE SOLVED: To provide a transmitter capable of directivity control after converting an analog output signal of a microphone into a digital signal of a pulse train through oversampling low-amplitude encoding and improving noise tolerance without increasing the number of wiring lines. SOLUTION: The output signal of the microphone is converted into an oversampled low-amplitude resolution encoded signal by the microphone, the directivity control is performed by performing delay and gain control over the digital signal consisting of the pulse train, and then the signal is converted into a PCM signal to perform audio digital processing. COPYRIGHT: (C)2007,JPO&INPIT
摘要翻译：要解决的问题：为了提供一种能够通过过采样低振幅编码将麦克风的模拟输出信号转换为脉冲序列的数字信号的发射机，并且在不增加布线数量的情况下提高噪声容限。解决方案：麦克风的输出信号由麦克风转换为过采样的低幅度分辨率编码信号，通过对由脉冲串组成的数字信号执行延迟和增益控制来执行方向性控制，然后信号被转换成PCM信号以执行音频数字处理。版权所有（C）2007，JPO＆INPIT

16. 发明专利

JP2007081814A Waveform data form discriminating method, signal coding method, and apparatus, program, and recording medium using the methods 有权
标题翻译：波形数据形式识别方法，信号编码方法以及使用该方法的装置，程序和记录介质
公开(公告)号：JP2007081814A
公开(公告)日：2007-03-29
申请号：JP2005266825
申请日：2005-09-14
申请人： Nippon Telegr & Teleph Corp , 日本電信電話株式会社
发明人： MORIYA TAKEHIRO , HARADA NOBORU , KAMAMOTO MASARU
IPC分类号： H03M7/30 , G10L11/00
摘要： PROBLEM TO BE SOLVED: To provide a technology of discriminating a waveform data form and efficiently coding waveform data even when part or all of a storage form of the waveform data is not recognized.
SOLUTION: A waveform data form discriminating method disclosed herein includes the steps of: assuming received digital waveform data to be sample value sequences of a plurality of presetting data forms when the data are recognized to be the waveform data but the storage form of the waveform data is unknown; discriminating likely data of the waveform data by using relationships between sample values of the sample value sequences; determining a data form discriminated to be most likely waveform data as the data form of the received digital waveform data; and coding all or part of the received digital data on the basis of a result of the data form discrimination.
COPYRIGHT: (C)2007,JPO&INPIT
摘要翻译：要解决的问题：即使在波形数据的存储形式的一部分或全部不被识别的情况下，提供区分波形数据形式并有效地编码波形数据的技术。解决方案：本文公开的波形数据形式鉴别方法包括以下步骤：当数据被识别为波形数据时，将接收到的数字波形数据作为多个预设数据形式的采样值序列，波形数据未知; 通过使用样本值序列的样本值之间的关系来区分波形数据的可能数据; 确定被识别为最可能的波形数据的数据形式作为所接收的数字波形数据的数据形式; 并根据数据形式辨别的结果对全部或部分接收到的数字数据进行编码。版权所有（C）2007，JPO＆INPIT

17. 发明专利

JP2007006337A Signal encoding apparatus, method and program, recording medium and signal codec method 有权
标题翻译：信号编码设备，方法和程序，记录介质和信号编解码方法
公开(公告)号：JP2007006337A
公开(公告)日：2007-01-11
申请号：JP2005186510
申请日：2005-06-27
申请人： Nippon Telegr & Teleph Corp , 日本電信電話株式会社
发明人： HARADA NOBORU , MORIYA TAKEHIRO
IPC分类号： H03M7/30 , H03M7/24
摘要： PROBLEM TO BE SOLVED: To separate a signal into an integer part and an error part so as to minimize a code amount.
SOLUTION: In the present invention, a shift amount determination section calculates a shift amount candidate so that a sample value of a maximum amplitude value within a current frame becomes the maximum amplitude that can be represented with an integer part while recording, in a frame shift amount holding buffer, a shift amount of at least one past frame, and the shift amount candidate and the shift amount recorded in said frame shift amount holding buffer are used to determine a shift amount of the current frame in accordance with a predetermined reference, so that a range of the amplitude of a signal to be encoded as an integer part is determined. Furthermore, a demultiplexer demultiplexes said digital signal into an integer signal and an error signal in accordance with the shift amount determined by said shift amount determination section, the integer signal is encoded by integer signal encoding, and the error signal is encoded by error signal encoding. An output of the integer signal encoding section and an output of the error signal encoding section are multiplexed by a multiplexer.
COPYRIGHT: (C)2007,JPO&INPIT
摘要翻译：要解决的问题：将信号分离成整数部分和错误部分，以使代码量最小化。解决方案：在本发明中，偏移量确定部分计算移位量候选，使得当前帧内的最大幅度值的采样值变为在记录时可以用整数部分表示的最大幅度，使用帧偏移量保持缓冲器，至少一个过去帧的移位量和移位量候选以及记录在所述帧偏移量保持缓冲器中的移位量，以根据预定的速度确定当前帧的移位量参考，使得要被编码为整数部分的信号的幅度的范围被确定。此外，解复用器根据由所述移位量确定部分确定的移位量将所述数字信号解复用为整数信号和误差信号，整数信号由整数信号编码编码，并且误差信号由误差信号编码。整数信号编码部分的输出和误差信号编码部分的输出由多路复用器复用。版权所有（C）2007，JPO＆INPIT

18. 发明专利

JP2004320237A Real-time processing device, real-time control method, program thereof, and information storage medium 有权
标题翻译：实时处理设备，实时控制方法，程序及信息存储介质
公开(公告)号：JP2004320237A
公开(公告)日：2004-11-11
申请号：JP2003109321
申请日：2003-04-14
申请人： Nippon Telegr & Teleph Corp , 日本電信電話株式会社
发明人： HARADA NOBORU
IPC分类号： H04L13/08
摘要： PROBLEM TO BE SOLVED: To provide a real-time processing device which is capable of carrying out real-time processing with a constant interval on average when the prescribed target processing is executed with a constant interval using a processing halt instruction that executes a prescribed target processing with a prescribed interval, when a time longer than a downtime set up by an instruction elapses after the instruction is set up, and to provide a real-time control method, programs thereof, and an information storage medium.
SOLUTION: A Sleep Time is set up taking a time difference ERR between a Target Time as a start target time of packet reading and an actual Start Time at which packet reading is actually started and a time required for actually reading out a packet into consideration, and an Sleep instruction is executed. When a following Target Time gets past, processing keeps continuing. Furthermore, when the number of packets in a buffer exceeds an allowable maximum value Nmax, the buffer is cleared, and delay recovery processing is reset.
COPYRIGHT: (C)2005,JPO&NCIPI
摘要翻译：要解决的问题：提供一种实时处理装置，当使用执行的处理暂停指令以规定的目标处理以一定的间隔执行时，能够以恒定的间隔平均执行实时处理当在指令建立之后经过指令的时间长于时间长于规定间隔的规定的目标处理，并提供其实时控制方法，程序和信息存储介质时。
解决方案：睡眠时间被设置为在分组读取的开始目标时间的目标时间和实际开始分组读取的实际开始时间之间的时间差ERR，以及实际读取分组所需的时间并且执行休眠指令。当以下目标时间过去时，处理继续。此外，当缓冲器中的分组数量超过允许的最大值Nmax时，缓冲器被清除，并且延迟恢复处理被重置。版权所有（C）2005，JPO＆NCIPI

19. 发明专利

JP2014222369A ＰＡＲＣＯＲ係数量子化方法、装置、プログラム有权
标题翻译： PARCOR系数量化方法与设备及程序
公开(公告)号：JP2014222369A
公开(公告)日：2014-11-27
申请号：JP2014168381
申请日：2014-08-21
申请人：日本電信電話株式会社 , Nippon Telegr & Teleph Corp
发明人： KAMAMOTO MASARU , HARADA NOBORU , MORIYA TAKEHIRO
IPC分类号： G10L19/06 , G10L19/00 , G10L19/13 , G10L19/16
CPC分类号： G10L19/06 , G10L19/0017
摘要：【課題】高圧縮ロスレス符号化のためのPARCOR係数量子化技術を提供する。【解決手段】PARCOR係数の量子化を行う際に、PARCOR係数が所定ビット数を持つ値として表されているとして、入力されたPARCOR係数の絶対値が大きいほど、最上位ビットから最下位ビットへ向かってより多くの有効ビットを含んで表される値を出力する。【選択図】図4
摘要翻译：要解决的问题：提供用于高压缩无损编码的PARCOR（部分自相关）系数量化技术。解决方案：当PARCOR系数被量化时，假设PARCOR系数被指示为具有预定位数的值作为绝对值输入的PARCOR系数的值较大，从最高有效位到最低有效位的更有效位表示的值被输出。

20. 发明专利

JP2014164091A Encoder, decoder, method of them, program, and recording medium 有权
标题翻译：编码器，解码器，它们的方法，程序和记录介质
公开(公告)号：JP2014164091A
公开(公告)日：2014-09-08
申请号：JP2013034554
申请日：2013-02-25
申请人： Nippon Telegr & Teleph Corp , 日本電信電話株式会社
发明人： MORIYA TAKEHIRO , KAMAMOTO MASARU , HARADA NOBORU , CHIBA HIRONOBU
IPC分类号： G10L19/08 , G10L19/09 , H03M7/30
摘要： PROBLEM TO BE SOLVED: To improve the encoding quality in encoding and decoding using a pitch sharpening technology.SOLUTION: A pitch sharpening coefficient is determined on the basis of the pitch cycle difference between the time section as a processing object and the time section before the processing object, the estimated value of the possibility of voiced sound time section, or the predicted gain. A cycled pulse sequence by applying the cycling corresponding to the pitch sharpening coefficient to one or more pulse sequences from a fixed codebook.
摘要翻译：要解决的问题：使用俯仰锐化技术来提高编码和解码中的编码质量。解决方案：基于作为处理对象的时间段和时间段之间的音调周期差异来确定音高锐化系数处理对象，浊音时间段的可能性的估计值，或预测的增益。循环脉冲序列通过将对应于音调锐化系数的循环应用于来自固定码本的一个或多个脉冲序列。

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