公开(公告)号：US20060004569A1

公开(公告)日：2006-01-05

申请号：US11165695

申请日：2005-06-24

申请人： Yasuo Yoshioka ,  Alex Loscos

发明人： Yasuo Yoshioka ,  Alex Loscos

IPC分类号： G10L19/14

CPC分类号： G10L13/033 ,  G10L2021/0135

摘要： Envelope identification section generates input envelope data (DEVin) indicative of a spectral envelope (EVin) of an input voice. Template acquisition section reads out, from a storage section, converting spectrum data (DSPt) indicative of a frequency spectrum (SPt) of a converting voice. On the basis of the input envelope data (DEVin) and the converting spectrum data (DSPt), a data generation section specifies a frequency spectrum (SPnew) corresponding in shape to the frequency spectrum (SPt) of the converting voice and having a substantially same spectral envelope as the spectral envelope (EVin) of the input voice, and the data generation section generates new spectrum data (DSPnew) indicative of the frequency spectrum (SPnew). Reverse FFT section and output processing section generates an output voice signal (Snew) on the basis of the new spectrum data (DSPnew).

摘要翻译： 信封识别部分生成表示输入声音的频谱包络（EVin）的输入包络数据（DEVin）。 模板获取部从存储部读出表示转换语音的频谱（SPt）的频谱数据（DSPt）。 基于输入包络数据（DEVin）和转换频谱数据（DSPt），数据生成部分指定与转换声音的频谱（SPt）形状对应的频谱（SPnew），并具有基本相同 频谱包络作为输入语音的频谱包络（EVin），并且数据产生部分生成指示频谱（SPnew）的新频谱数据（DSPnew）。 反向FFT部分和输出处理部分基于新的频谱数据（DSPnew）生成输出语音信号（Snew）。

公开(公告)号：US08117031B2

公开(公告)日：2012-02-14

申请号：US11961580

申请日：2007-12-20

申请人： Takehiko Kawahara ,  Yasuo Yoshioka

发明人： Takehiko Kawahara ,  Yasuo Yoshioka

IPC分类号： G10L17/00

CPC分类号： G10L25/48 ,  G10L17/04 ,  G10L2015/0631 ,  G10L2015/0635

摘要： A voice processing apparatus has a storage device that stores registration information containing a characteristic parameter of a given voice. The voice processing apparatus is further provided with a judgment unit, a management unit and a notification unit. The judgment unit judges whether an input voice is appropriate or not for creating or updating the registration information based on a degree of a difference between an inter-band correlation matrix of an input voice acquired this time and an inter-band correlation matrix of another input voice that is judged as being appropriate last time. The management unit creates or updates the registration information based on a characteristic parameter of the input voice when the judgment unit judges that the input voice is appropriate. The notification unit notifies a speaker of the input voice when the judgment unit judges that the input voice is inappropriate.

摘要翻译： 语音处理装置具有存储包含给定语音的特征参数的登记信息的存储装置。 语音处理装置还具有判断单元，管理单元和通知单元。 判断单元基于本次获取的输入语音的带间相关矩阵与另一输入的频带间相关矩阵之间的差异程度来判断输入语音是否适合于创建或更新注册信息 上次被判断为合适的声音。 当判断单元判断输入的声音是适当的时，管理单元基于输入声音的特性参数来创建或更新注册信息。 当判断单元判断输入的声音不合适时，通知单元向扬声器通知输入的声音。

公开(公告)号：US08073688B2

公开(公告)日：2011-12-06

申请号：US11165695

申请日：2005-06-24

申请人： Yasuo Yoshioka ,  Alex Loscos

发明人： Yasuo Yoshioka ,  Alex Loscos

IPC分类号： G10L19/14

CPC分类号： G10L13/033 ,  G10L2021/0135

摘要： Envelope identification section generates input envelope data (DEVin) indicative of a spectral envelope (EVin) of an input voice. Template acquisition section reads out, from a storage section, converting spectrum data (DSPt) indicative of a frequency spectrum (SPt) of a converting voice. On the basis of the input envelope data (DEVin) and the converting spectrum data (DSPt), a data generation section specifies a frequency spectrum (SPnew) corresponding in shape to the frequency spectrum (SPt) of the converting voice and having a substantially same spectral envelope as the spectral envelope (EVin) of the input voice, and the data generation section generates new spectrum data (DSPnew) indicative of the frequency spectrum (SPnew). Reverse FFT section and output processing section generates an output voice signal (Snew) on the basis of the new spectrum data (DSPnew).

摘要翻译： 信封识别部分生成表示输入声音的频谱包络（EVin）的输入包络数据（DEVin）。 模板获取部从存储部读出表示转换语音的频谱（SPt）的频谱数据（DSPt）。 基于输入包络数据（DEVin）和转换频谱数据（DSPt），数据生成部分指定与转换声音的频谱（SPt）形状对应的频谱（SPnew），并具有基本相同 频谱包络作为输入语音的频谱包络（EVin），并且数据产生部分生成指示频谱（SPnew）的新频谱数据（DSPnew）。 反向FFT部分和输出处理部分基于新的频谱数据（DSPnew）生成输出语音信号（Snew）。

公开(公告)号：US07606709B2

公开(公告)日：2009-10-20

申请号：US10282536

申请日：2002-10-29

申请人： Yasuo Yoshioka ,  Hiraku Kayama ,  Xavier Serra ,  Jordi Bonada

发明人： Yasuo Yoshioka ,  Hiraku Kayama ,  Xavier Serra ,  Jordi Bonada

IPC分类号： G10L13/00 ,  G10L21/00

CPC分类号： G10L13/033 ,  G10L19/093 ,  G10L21/02 ,  G10L25/93 ,  G10L2021/0135

摘要： An apparatus is constructed for converting an input voice signal into an output voice signal according to a target voice signal. In the apparatus, an input device provides the input voice signal composed of original sinusoidal components and original residual components other than the original sinusoidal components. An extracting device extracts original attribute data from at least the sinusoidal components of the input voice signal. The original attribute data is characteristic of the input voice signal. A synthesizing device synthesizes new attribute data based on both of the original attribute data derived from the input voice signal and target attribute data being characteristic of the target voice signal composed of target sinusoidal components and target residual components other than the sinusoidal components. The target attribute data is derived from at least the target sinusoidal components. An output device operates based on the new attribute data and either of the original residual component and the target residual component for producing the output voice signal.

摘要翻译： 一种根据目标语音信号将输入语音信号转换为输出语音信号的装置。 在该装置中，输入装置提供由原始正弦分量和原始正弦分量以外的原始剩余分量组成的输入语音信号。 提取装置从至少输入语音信号的正弦分量中提取原始属性数据。 原始属性数据是输入语音信号的特征。 合成装置基于从输入语音信号导出的原始属性数据和由目标正弦分量和除了正弦分量之外的目标残差分量组成的目标语音信号的特征的目标属性数据，合成新的属性数据。 目标属性数据至少从目标正弦分量导出。 输出装置基于新的属性数据和原始剩余分量中的任一个和用于产生输出语音信号的目标剩余分量进行操作。

公开(公告)号：US20090063146A1

公开(公告)日：2009-03-05

申请号：US12198232

申请日：2008-08-26

申请人： Yasuo Yoshioka

发明人： Yasuo Yoshioka

IPC分类号： G10L17/00 ,  G10L21/00

CPC分类号： G10L17/26 ,  G10L25/90 ,  G10L25/93

摘要： In a voice processing device, a male voice index calculator calculates a male voice index indicating a similarity of the input sound relative to a male speaker sound model. A female voice index calculator calculates a female voice index indicating a similarity of the input sound relative to a female speaker sound model. A first discriminator discriminates the input sound between a non-human-voice sound and a human voice sound which may be either of the male voice sound or the female voice sound. A second discriminator discriminates the input sound between the male voice sound and the female voice sound based on the male voice index and the female voice index in case that the first discriminator discriminates the human voice sound.

摘要翻译： 在语音处理装置中，男性声音指标计算器计算表示输入声音相对于男性扬声器声音模型的相似度的男性声音指数。 女性声音指标计算器计算表示输入声音相对于女性扬声器声音模型的相似度的女性声音指数。 第一鉴别器鉴别可能是男性声音或女性声音之一的非人声音和人声音之间的输入声音。 在第一鉴别器识别人声音的情况下，第二鉴别器基于男性声音索引和女性声音索引来鉴别男性声音和女性声音之间的输入声音。

公开(公告)号：US20080192954A1

公开(公告)日：2008-08-14

申请号：US11886044

申请日：2006-03-10

申请人： Yoshikazu Honji ,  Yasuo Yoshioka ,  Tetsu Kobayashi ,  Akio Takahashi

发明人： Yoshikazu Honji ,  Yasuo Yoshioka ,  Tetsu Kobayashi ,  Akio Takahashi

IPC分类号： H04B1/00

CPC分类号： G10K15/04

摘要： Microphones are provided at an air inlet of the engine and a vehicle-cabin-side wall surface of an engine room, and engine sounds are picked up. The engine sound is processed by a signal processing section, and the processed engine sound is output from a speaker provided in a vehicle cabin. The signal processing section is provided with a filter which simulates a sound insulation characteristic of the vehicle cabin and a transformation section for processing the engine sound according to driving condition. A spectrum transformation characteristic of the transformation section is determined according to values detected by a vehicle speed sensor, an engine speed sensor, and an accelerator depression sensor, and a spectrum of the engine sound is transformed by means of specification of the spectrum transformation characteristic, thereby enhancing an engine sound.

摘要翻译： 在发动机的进气口和发动机室的车厢侧壁面设置有麦克风，并且拾取发动机声音。 发动机声音由信号处理部分处理，并且处理的发动机声音从设置在车厢中的扬声器输出。 信号处理部设置有模拟车厢的隔音特性的过滤器和根据驾驶状况处理发动机声音的变换部。 根据由车速传感器，发动机转速传感器和加速器凹陷传感器检测的值来确定变换部的频谱变换特性，通过频谱变换特性的规格来变换发动机声音的频谱， 从而增强发动机声音。

公开(公告)号：US07135636B2

公开(公告)日：2006-11-14

申请号：US10375272

申请日：2003-02-27

申请人： Hideki Kemmochi ,  Yasuo Yoshioka ,  Jordi Bonada

发明人： Hideki Kemmochi ,  Yasuo Yoshioka ,  Jordi Bonada

IPC分类号： G10H1/06 ,  G10H7/00

CPC分类号： G10L13/00 ,  G10H7/00 ,  G10H2240/056 ,  G10H2250/235 ,  G10H2250/455

摘要： A method for synthesizing a natural-sounding singing voice divides performance data into a transition part and a long sound part. The transition part is represented by articulation (phonemic chain) data that is read from an articulation template database and is outputted without modification. For the long sound part, a new characteristic parameter is generated by linearly interpolating characteristic parameters of the transition parts positioned before and after the long sound part and adding thereto a changing component of stationary data that is read from a constant part (stationary) template database. An associated apparatus for carrying out the singing voice synthesizing method includes a phoneme database for storing articulation data for the transition part and stationary data for the long sound part, a first device for outputting the articulation data, and a second device for outputting the newly-generated characteristic parameter of the long sound part.

摘要翻译： 用于合成自然发声的歌声的方法将演奏数据分成转换部分和长音部分。 过渡部分由从关节运动模板数据库读取并且没有修改地输出的关节（音素链）数据表示。 对于长音部分，通过线性内插位于长声部分之前和之后的过渡部分的特征参数，并且向其添加从恒定部分（静止）模板数据库读取的静止数据的变化分量，生成新的特征参数 。 用于执行歌唱声合成方法的相关装置包括用于存储用于转换部分的发音数据的音素数据库和用于长音部分的固定数据，用于输出关节数据的第一装置，以及用于输出新音符的第二装置， 生成长音部分的特征参数。

公开(公告)号：US20050049875A1

公开(公告)日：2005-03-03

申请号：US10951328

申请日：2004-09-27

申请人： Takahiro Kawashima ,  Yasuo Yoshioka ,  Pedro Cano ,  Alex Loscos ,  Xavier Serra ,  Mark Schiementz ,  Jordi Bonada

发明人： Takahiro Kawashima ,  Yasuo Yoshioka ,  Pedro Cano ,  Alex Loscos ,  Xavier Serra ,  Mark Schiementz ,  Jordi Bonada

IPC分类号： G10L13/02 ,  G10L21/00 ,  G10L13/00

CPC分类号： G10L13/033 ,  G10L2021/0135

摘要： A voice converting apparatus is constructed for converting an input voice into an output voice according to a target voice. In the apparatus, a storage section provisionally stores source data, which is associated to and extracted from the target voice. An analyzing section analyzes the input voice to extract therefrom a series of input data frames representing the input voice. A producing section produces a series of target data frames representing the target voice based on the source data, while aligning the target data frames with the input data frames to secure synchronization between the target data frames and the input data frames. A synthesizing section synthesizes the output voice according to the target data frames and the input data frames. In the recognizing feature analysis, a characteristic analyzer extracts from the input voice a characteristic vector. A memory memorizes target behavior data representing a behavior of the target voice. An alignment processor determines a temporal relation between the input data frames and the target data frames according to the characteristic vector and the target behavior data so as to output alignment data. A target decoder produces the target data frames according to the alignment data, the input data frames and the source data containing phoneme of the target voice.

摘要翻译： 构成语音转换装置，用于根据目标语音将输入语音转换为输出语音。 在装置中，存储部临时存储与目标语音相关联并从其中提取的源数据。 分析部分分析输入声音以从中提取代表输入声音的一系列输入数据帧。 产生部分基于源数据产生一系列表示目标语音的目标数据帧，同时使目标数据帧与输入数据帧对齐，以确保目标数据帧与输入数据帧之间的同步。 合成部根据目标数据帧和输入数据帧合成输出声音。 在识别特征分析中，特征分析器从输入语音中提取特征向量。 存储器存储表示目标语音行为的目标行为数据。 对准处理器根据特征向量和目标行为数据确定输入数据帧和目标数据帧之间的时间关系，以输出对准数据。 目标解码器根据对准数据，输入数据帧和包含目标声音的音素的源数据产生目标数据帧。

公开(公告)号：US08315855B2

公开(公告)日：2012-11-20

申请号：US12460650

申请日：2009-07-22

申请人： Yasuo Yoshioka

发明人： Yasuo Yoshioka

IPC分类号： G10L11/04

CPC分类号： G10L21/003 ,  G10H1/366 ,  G10L13/033

摘要： Character extraction section extracts character amounts, pertaining to a prosody of voice, from a voice signal sequentially in a time-serial manner. Difference value calculation calculates a difference value between each of the extracted character amounts and a reference value. Processing values, corresponding to the individual character amounts, are generated in accordance with the respective difference values, and a voice processing section controls the individual character amounts of the voice signal in accordance with the processing values corresponding to the character amounts and thereby generates an output signal having a prosody changed from the prosody of the voice signal.

摘要翻译： 字符提取部分以语音信号顺序地以时间序列方式提取与声音韵律有关的字符量。 差分值计算计算每个提取的字符量和参考值之间的差值。 根据各自的差值生成与各个字符量对应的处理值，声音处理部根据与字符量对应的处理值来控制语音信号的个别字符量，从而生成输出 具有从语音信号的韵律改变的韵律的信号。

公开(公告)号：US08300834B2

公开(公告)日：2012-10-30

申请号：US11916993

申请日：2006-06-28

申请人： Yasuo Yoshioka

发明人： Yasuo Yoshioka

IPC分类号： H04R29/00

CPC分类号： G10L25/78 ,  G10L25/90

摘要： Even in a state that the change of an environmental noise cannot be anticipated, a sound generating period in an audio signal can be specified with high accuracy. Sound in an audio space in which an audio signal processing system 1 is disposed is always collected by a microphone 20 and inputted to an audio signal processing device 10 as an audio signal. Before a user carried out a prescribed operation, the audio signals inputted from the microphone 20 are sequentially stored in a first buffer 121. After the prescribed operation is carried out, the audio signals are sequentially stored in a second buffer 122. A specifying part 114 considers the level of the audio signal stored in the first buffer 121 as the level of the environmental noise and the level of the audio signal sequentially stored in the second buffer 122 as the level of sound generated at a current time to calculate an S/N ratio. The specifying part 114 sequentially decides whether or not the calculated S/N ratio satisfies a prescribed condition to specify the sound generating period in the audio signal.

摘要翻译： 即使在不能预期到环境噪声的变化的情况下，也可以高精度地规定音频信号的声音发生期间。 音频信号处理系统1设置在音频空间中的声音总是由麦克风20收集，并作为音频信号输入到音频信号处理装置10。 在用户执行规定的操作之前，从麦克风20输入的音频信号被顺序地存储在第一缓冲器121中。在执行规定的操作之后，音频信号被顺序地存储在第二缓冲器122中。指定部分114 将存储在第一缓冲器121中的音频信号的电平视为环境噪声的电平和顺序存储在第二缓冲器122中的音频信号的电平，作为当前时间产生的声音电平，以计算S / N 比。 指定部分114顺序地确定所计算的S / N比是否满足规定条件以指定音频信号中的声音产生时段。