公开(公告)号：KR100348899B1

公开(公告)日：2002-08-14

申请号：KR1020000054960

申请日：2000-09-19

Applicant: 한국전자통신연구원

Inventor： 김형중 ,  이인성 ,  김종학 ,  박만호 ,  윤병식 ,  최송인 ,  김대식

IPC: G10L19/04

Abstract: 본 발명은 하모닉 모델을 사용하는 유/무성음 혼합신호의 하모닉 노이즈 음성 부호화기 및 부호화 방법에 관한 것으로서, 입력되는 LPC 잔여신호를 캡스트럼을 이용하여 무성음 성분인 노이즈를 분리한 후 LPC 분석법으로 스펙트럴을 예측하여 상기 노이즈를 부호화 하는 노이즈-스펙트럴 추정 수단을 포함하는 것을 특징으로 하며, 유/무성음 혼합 신호를 기존의 하모닉 모델에 캡스트럼-LPC 분석법을 통해 예측된 노이즈 스펙트럴 모델을 사용하여 효과적으로 노이즈 분석을 하여 부호화 함으로써, 보다 개선된 음질을 구현할 수 있는 것을 특징으로 한다.

公开(公告)号：KR100383668B1

公开(公告)日：2003-05-14

申请号：KR1020000054959

申请日：2000-09-19

Applicant: 한국전자통신연구원

Inventor： 김형중 ,  이인성 ,  김종학 ,  박만호 ,  윤병식 ,  최송인 ,  김대식

IPC: G10L19/02

CPC classification number: G10L19/02

Abstract: The present invention relates to a time-separated speech coder, that codes the transitional signal of voiced/unvoiced sound through harmonic speech coding. The time-separated speech coder comprises a transitional excitation signal analysis/synthesis means 10 for coding the transitional signal by extracting the harmonic model parameters of both transitional analyzers after detecting transitional point and generating sinusoidal waveforms according to variable transitional point separating both transitional analyzers. By the transitional point at which energy varies abruptly and the time-separated coding based on the transitional point, more improved speech quality than in the general harmonic speech coder can be obtained using the time-separated speech coder by increasing the representation capability of the transitional signal with large energy variation, after being adapted it to the variable transitional point.

Abstract translation: 本发明涉及一种时间分离的语音编码器，它通过谐波语音编码来编码有声/无声声音的过渡信号。 时间分离语音编码器包括过渡激励信号分析/合成装置10，用于通过在检测到过渡点之后提取两个过渡分析器的谐波模型参数并根据分离两个过渡分析器的可变过渡点产生正弦波形来对过渡信号进行编码。 通过能量突然变化的过渡点和基于过渡点的时间分隔编码，通过使用时间分离的语音编码器，通过增加过渡的表示能力，可以获得比一般谐波语音编码器更多改进的语音质量 信号经过适应变化的过渡点后，具有较大的能量变化。

公开(公告)号：KR1020020022257A

公开(公告)日：2002-03-27

申请号：KR1020000054960

申请日：2000-09-19

Applicant: 한국전자통신연구원

Inventor： 김형중 ,  이인성 ,  김종학 ,  박만호 ,  윤병식 ,  최송인 ,  김대식

IPC: G10L19/04

CPC classification number: G10L19/08 ,  G10L2025/935

Abstract: PURPOSE: A harmonic-noise speech coding system and method using cepstrum analysis are provided to enable effective coding through harmonic-noise speech coding algorithm that estimates noise spectral component to offer improved sound quality. CONSTITUTION: A harmonic-noise speech coding system(100) includes a noise-spectral estimator that separates noise from an input LPC residual signal using cepstrum and estimates spectral through LCD analysis to code the noise. The noise-spectral estimator includes a log value extracting means for extracting negative log value spectrum of the cepstrum extracted through the cepstrum analysis, and an amplitude extracting means for extracting a valley portion of mixed signal spectral, corresponding to the negative log value spectrum region. The noise-spectral estimator further has an LPC analyzer for extracting spectral information of the extracted noise spectral, an LSP converter for converting extracted LPC parameters into LSP parameters, and a gain calculator for calculating the gain of noise component.

Abstract translation: 目的：提供使用倒谱分析的谐波噪声语音编码系统和方法，以通过估计噪声频谱分量的谐波噪声语音编码算法实现有效编码，从而提高音质。 构成：谐波噪声语音编码系统（100）包括噪声频谱估计器，其使用倒频谱将噪声与输入LPC残差信号分离，并通过LCD分析估计频谱以对噪声进行编码。 噪声谱估计器包括用于提取通过倒频谱分析提取的倒频谱的负对数值谱的对数值提取装置，以及用于提取与负对数值谱区对应的混合信号谱的谷部的振幅提取装置。 噪声频谱估计器还具有用于提取所提取的噪声频谱的频谱信息的LPC分析器，用于将提取的LPC参数转换为LSP参数的LSP转换器，以及用于计算噪声分量增益的增益计算器。

公开(公告)号：KR1020020022256A

公开(公告)日：2002-03-27

申请号：KR1020000054959

申请日：2000-09-19

Applicant: 한국전자통신연구원

Inventor： 김형중 ,  이인성 ,  김종학 ,  박만호 ,  윤병식 ,  최송인 ,  김대식

IPC: G10L19/02

CPC classification number: G10L19/02

Abstract: PURPOSE: A system and a method for coding speech using time division coding algorithm are provided to code speech by detecting the transition time of a transition section and dividing the section in order to obtain improved speech quality of the transition section. CONSTITUTION: A time division speech coder for coding a voiced/voiceless sound transition signal through harmonic speech coding includes an excitation signal transition section analyzer(10). The excitation signal transition section analyzer has a transition time detector, a harmonic excitation signal analyzer, and a harmonic excitation signal adder. The transition time detector detects transition time for grasping the transition section of the transition signal. The harmonic excitation signal analyzer extracts a harmonic model parameter of the detected transition section. The excitation signal transition section adder adds the harmonic model parameter.

Abstract translation: 目的：提供一种使用时分编码算法对语音进行编码的系统和方法，通过检测转换部分的转换时间并对该部分进行划分，以获得改进的转换部分的语音质量来编码语音。 构成：用于通过谐波语音编码对浊音/无声声音转换信号进行编码的时分语音编码器包括激励信号转换部分析器（10）。 励磁信号过渡段分析仪具有过渡时间检测器，谐波激励信号分析仪和谐波激励信号加法器。 转移时间检测器检测用于掌握转换信号的转换部分的转换时间。 谐波激励信号分析仪提取检测到的过渡段的谐波模型参数。 激励信号转换部分加法器加上谐波模型参数。