公开(公告)号：EP2237267A4

公开(公告)日：2012-01-18

申请号：EP08863826

申请日：2008-12-22

Applicant: PANASONIC CORP

Inventor： MORII TOSHIYUKI

IPC: G10L19/00 ,  G10L19/008

CPC classification number: G10L19/008

公开(公告)号：EP2120234A4

公开(公告)日：2011-08-03

申请号：EP08710503

申请日：2008-02-29

Applicant: PANASONIC CORP

Inventor： MORII TOSHIYUKI ,  OSHIKIRI MASAHIRO ,  YAMANASHI TOMOFUMI

IPC: G10L19/10 ,  G10L19/02 ,  G10L19/032 ,  G10L19/08

CPC classification number: G10L19/032 ,  G10L19/0212 ,  G10L19/10

Abstract: Provided is an encoding device which can reduce the encoding distortion as compared to the conventional technique and can obtain a preferable sound quality for auditory sense. In the encoding device, a shape quantization unit (111) quantizes the shape of an input spectrum with a small number of pulse positions and polarities. The shape quantization unit (111) sets a pulse amplitude width to be searched later upon search of the pulse position to a value not greater than the pulse amplitude width which has been searched previously. A gain quantization unit (112) calculates a gain of a pulse searched by the shape quantization unit (111) for each of bands.

Abstract translation: 提供了一种与传统技术相比可以减少编码失真并且可以获得用于听觉的优选声音质量的编码装置。 在编码装置中，形状量化单元（111）以少量的脉冲位置和极性量化输入频谱的形状。 形状量化单元（111）将脉冲位置的脉冲幅度宽度设定为不大于之前已经搜索过的脉冲幅度宽度的值。 增益量化单元（112）针对每个频带计算由形状量化单元（111）搜索的脉冲的增益。

公开(公告)号：EP2099026A4

公开(公告)日：2011-02-23

申请号：EP07850564

申请日：2007-12-13

Applicant: PANASONIC CORP

Inventor： MORII TOSHIYUKI

IPC: G10L19/14 ,  G10L19/12 ,  G10L19/125 ,  G10L19/26

CPC classification number: G10L19/26 ,  G10L19/125

公开(公告)号：EP1881488A4

公开(公告)日：2008-12-10

申请号：EP06745821

申请日：2006-04-28

Applicant: PANASONIC CORP

Inventor： SATOH KAORU ,  MORII TOSHIYUKI ,  YAMANASHI TOMOFUMI

IPC: G10L19/14

CPC classification number: G10L19/24

Abstract: An encoder generating a decoded signal with an improved quality by scalable encoding by canceling the characteristic inherent to the encoder and causing degradation of quality of the decoded signal. In the encoder, a first encoding section (102) encodes the input signal after down sampling, a first decoding section (103) decodes first encoded information outputted from the first encoding section (102), an adjusting section (105) adjusts the first decoded signal after up sampling by convoluting the first decoded signal after up sampling and an impulse response for adjustment, an adder (107) inverses the polarity of adjusted first decoded signal and adds the first decoded signal having the inverted polarity to the input signal, a second encoding section (108) encodes the residual signal outputted from the adder (107), and a multiplexing section (109) multiplexes the first encoded information outputted from the first encoding section (102) and the second encoded information outputted from the second encoding section (108).

公开(公告)号：JP2013101212A

公开(公告)日：2013-05-23

申请号：JP2011244490

申请日：2011-11-08

Applicant: Panasonic Corp ,  パナソニック株式会社

Inventor： MORII TOSHIYUKI

IPC: G10L25/90 ,  G10L19/12 ,  G10L25/06 ,  H03M7/30

Abstract: PROBLEM TO BE SOLVED: To perform highly accurate pitch analysis with a small calculation amount even to a target signal with a high sampling rate.SOLUTION: This pitch analysis device for analyzing a pitch of a target signal mainly for signal analysis or signal encoding includes: a low pass filter 302 for limiting a frequency band of a target signal 301, and for outputting a transmission target signal whose frequency has been adjusted; a thinned-out correlation value calculation part 303 for extracting the signal value of a transmission target signal with a second sampling interval which is longer than a first sampling interval as the sampling interval of the target signal, and for calculating the auto-correction of the transmission target signal and the power of the transmission target signal; and a pitch determination part 304 for determining the pitch of the target signal on the basis of the calculated auto-correction and power.

Abstract translation: 要解决的问题：即使对于具有高采样率的目标信号，也以小的计算量执行高精度的音调分析。 解决方案：用于分析主要用于信号分析或信号编码的目标信号的音高的音调分析装置包括：低通滤波器302，用于限制目标信号301的频带，并且用于输出发送目标信号， 频率已调整; 减去相关值计算部303，用于以比目标信号的采样间隔长的第二采样间隔提取发送目标信号的信号值，并且计算自动校正 传输目标信号和传输目标信号的功率; 以及基于所计算的自动校正和功率来确定目标信号的音高的音调确定部304。 版权所有（C）2013，JPO＆INPIT

公开(公告)号：JP2013068847A

公开(公告)日：2013-04-18

申请号：JP2011208096

申请日：2011-09-22

Applicant: Panasonic Corp ,  パナソニック株式会社

Inventor： MORII TOSHIYUKI

IPC: G10L19/08 ,  G10L19/083 ,  G10L19/10

Abstract: PROBLEM TO BE SOLVED: To achieve high coding performance with relatively low computational complexity by performing searches for positions and polarities of pulses and quantization of gains in many stages to distribute proper amounts of bits for the positions and polarities as well as quantization of gains.SOLUTION: In position and polarity search processing ST21 in the first stage, positions and polarities of pulses in the case that gains of two pulses out of four pulses are ideal gains are searched. In gain VQ processing ST22 in the first stage, gains of pulses of which the positions and polarities have been searched in the position and polarity search processing ST21 in the first stage are quantized. In position and polarity search processing ST21 in the second stage, positions and polarities of pulses in the case that gains of two remaining pulses are ideal gains are searched. In gain VQ processing ST22 in the second stage, gains of pulses of which the positions and polarities have been searched in the position and polarity search processing ST21 in the second stage are quantized.

Abstract translation: 要解决的问题：为了通过执行脉冲的位置和极性的搜索以及许多级的增益的量化来实现具有相对低的计算复杂度的高编码性能，以分配用于位置和极性的适当的位数量以及量化 收益。

解决方案：在第一级的位置和极性搜索处理ST21中，搜索在四个脉冲中的两个脉冲的增益是理想增益的情况下的脉冲的位置和极性。 在第一级的增益VQ处理ST22中，量化在第一级中的位置和极性搜索处理ST21中搜索了位置和极性的脉冲的增益。 在第二级的位置和极性搜索处理ST21中，搜索在两个剩余脉冲的增益是理想增益的情况下的脉冲的位置和极性。 在第二级的增益VQ处理ST22中，对第二级的位置和极性搜索处理ST21中的位置和极性的脉冲的增益进行量化。 版权所有（C）2013，JPO＆INPIT

公开(公告)号：JP2013055417A

公开(公告)日：2013-03-21

申请号：JP2011190860

申请日：2011-09-01

Applicant: Panasonic Corp ,  パナソニック株式会社

Inventor： MORII TOSHIYUKI

IPC: H03M7/30 ,  G10L19/16

Abstract: PROBLEM TO BE SOLVED: To implement quantization performance matching bit number and complexity with increasing bit number, and reduce a final quantization distortion by preventing a later stage from increasing a quantization distortion once reduced in an earlier stage during multistage quantization.SOLUTION: VQ sections 301, 305, 310, 315 quantize a signal to be quantized to acquire a decoded signal. Subtraction sections 302, 307, 312 subtract the decoded signal from the signal to be quantized to produce an error component which is acquired as the signal to be quantized in the following stage. Determination sections 303, 308, 313 finish the quantization if the magnitude of the error component is less than a predetermined value. A code accumulation section 317 stores codes of respective stages of quantization, and when the quantization is finished, collectively outputs the stored codes of respective stages.

Abstract translation: 要解决的问题：通过增加比特数来实现量化性能匹配比特数和复杂度，并且通过防止稍后阶段在多级量化期间在较早阶段减少量化失真来减少最终量化失真。 解决方案：VQ部分301,305,310,315量化要量化的信号以获取解码信号。 减法部302,307,312从要量化的信号中减去解码信号，以产生作为在下一阶段中要量化的信号获取的误差分量。 如果误差分量的幅度小于预定值，则确定部分303,308,313完成量化。 代码累积部分317存储各级的代码，并且当量化完成时，共同地输出存储的各个级的代码。 版权所有（C）2013，JPO＆INPIT

公开(公告)号：JP2013152413A

公开(公告)日：2013-08-08

申请号：JP2012018235

申请日：2012-01-31

Applicant: Panasonic Corp ,  パナソニック株式会社

Inventor： MORII TOSHIYUKI

IPC: G10L25/21 ,  G10L25/06 ,  G10L25/18

Abstract: PROBLEM TO BE SOLVED: To reduce a calculation amount when band power analysis is performed.SOLUTION: In a band power analyzer 100 which analyzes power for every plurality of bands of input signals, an autocorrelation calculation part 101 calculates finite degrees of autocorrelation by using the input signals, a fixed coefficient storage part 103 stores coefficients for every plurality of bands, which are obtained by adding a plurality of cosine coefficients calculated for every frequency to every plurality of bands, and a band power calculation part 102 calculates power for every plurality of bands by performing a product-sum operation between the finite degrees of autocorrelation and the coefficients of the bands in the respective bands of the plurality of bands.

Abstract translation: 要解决的问题：为了减少进行频带功率分析时的计算量。解决方案：自动相关计算部101在对每个输入信号的多个频带进行功率分析的频带功率分析器100中，通过使用自相关计算部101来计算有限度的自相关 输入信号，固定系数存储部103存储通过将针对每个频率计算的多个余弦系数相加而获得的每多个频带的系数，并且频带功率计算部102计算每个频带的功率 通过在有限自适应度与多个频带的各个频带中的频带的系数之间进行乘积和运算来进行频带。

公开(公告)号：JP2009042734A

公开(公告)日：2009-02-26

申请号：JP2008045259

申请日：2008-02-26

Applicant: Panasonic Corp ,  パナソニック株式会社

Inventor： OSHIKIRI MASAHIRO ,  MORII TOSHIYUKI ,  YAMANASHI TOMOHITO

IPC: G10L19/02 ,  G10L19/083 ,  G10L19/16 ,  H03M7/30

CPC classification number: G10L19/00 ,  G10L19/005 ,  G10L19/02 ,  G10L19/0208 ,  G10L19/038 ,  G10L19/083 ,  G10L19/24 ,  G10L25/18

Abstract: PROBLEM TO BE SOLVED: To provide a voice encoding device for accurately encoding the shape of a spectrum of a signal with high tonality such as a vowel. SOLUTION: A subband composition part 151 divides a first layer error conversion coefficient to be encoded into M subbands to generate M subband conversion coefficients. A shape vector encoding part 152 encodes each of the M subband conversion coefficients to obtain M pieces of shape encoding information, and calculates each target gain of the M subband conversion coefficients. A gain vector composition part 153 composes one gain vector by using M target gains. A gain vector encoding part 154 obtains gain encoding information by encoding the gain vector. A multiplexing part 155 multiplexes shape encoding information with gain encoding information. COPYRIGHT: (C)2009,JPO&INPIT

Abstract translation: 要解决的问题：提供一种语音编码装置，用于对诸如元音的高音调的信号的频谱的形状进行精确编码。 解决方案：子带合成部分151将要编码的第一层错误转换系数划分为M个子带，以产生M个子带转换系数。 形状矢量编码部分152对M个子带转换系数中的每一个编码以获得M个形状编码信息，并且计算M个子带转换系数的每个目标增益。 增益矢量合成部153通过使用M个目标增益来构成一个增益矢量。 增益矢量编码部154通过对增益矢量进行编码而获得增益编码信息。 复用部155将形状编码信息与增益编码信息进行复用。 版权所有（C）2009，JPO＆INPIT

公开(公告)号：JP2013057792A

公开(公告)日：2013-03-28

申请号：JP2011195892

申请日：2011-09-08

Applicant: Panasonic Corp ,  パナソニック株式会社

Inventor： MORII TOSHIYUKI

IPC: G10L19/00 ,  G10L19/02

Abstract: PROBLEM TO BE SOLVED: To achieve an efficient use of information source codes at the time of coding time-series signals to suppress a calculation amount and improve coding performance in frequency-domain signals.SOLUTION: A first adjustment unit 102 performs a nonlinear amplitude adjustment to a first frequency spectrum. A CELP coding unit 104 generates a decoded signal by performing a CELP coding and a subsequent decoding. A second MDCT unit 105 subjects the decoded signal to an orthogonal transformation to generate a second frequency spectrum. A second adjustment unit 106 gives characteristics opposite to those which have been given to the first frequency spectrum during the amplitude adjustment in the first adjustment unit 102, to the second frequency spectrum. A subtraction unit 107 subtracts the second frequency spectrum to which the opposite characteristics have been given, from the first frequency spectrum to generate a residual spectrum. An FPC coding unit 108 codes the residual spectrum into an FPC code.

Abstract translation: 要解决的问题：为了在编码时间序列信号时有效利用信息源代码，以抑制计算量并提高频域信号中的编码性能。 解决方案：第一调整单元102对第一频谱进行非线性幅度调整。 CELP编码单元104通过执行CELP编码和随后的解码来生成解码信号。 第二MDCT单元105对经解码的信号进行正交变换以产生第二频谱。 第二调整单元106给出与在第一调整单元102中的幅度调整期间已经被赋予第一频谱的那些相反的特性到第二频谱。 减法单元107从第一频谱中减去具有相反特性的第二频谱，以产生残余频谱。 FPC编码单元108将残余频谱编码为FPC码。 版权所有（C）2013，JPO＆INPIT