公开(公告)号：US20110002473A1

公开(公告)日：2011-01-06

申请号：US12919694

申请日：2009-02-27

申请人： Tomohiro Nakatani ,  Takuya Yoshioka ,  Keisuke Kinoshita ,  Masato Miyoshi

发明人： Tomohiro Nakatani ,  Takuya Yoshioka ,  Keisuke Kinoshita ,  Masato Miyoshi

IPC分类号： H04B3/20

CPC分类号： G10L2021/02082

摘要： A sound source model storage section stores a sound source model that represents an audio signal emitted from a sound source in the form of a probability density function. An observation signal, which is obtained by collecting the audio signal, is converted into a plurality of frequency-specific observation signals each corresponding to one of a plurality of frequency bands. Then, a dereverberation filter corresponding to each frequency band is estimated by using the frequency-specific observation signal for the frequency band on the basis of the sound source model and a reverberation model that represents a relationship for each frequency band among the audio signal, the observation signal and the dereverberation filter. A frequency-specific target signal corresponding to each frequency band is determined by applying the dereverberation filter for the frequency band to the frequency-specific observation signal for the frequency band, and the resulting frequency-specific target signals are integrated.

摘要翻译： 声源模型存储部存储以概率密度函数的形式表示从声源发出的音频信号的声源模型。 通过收集音频信号而获得的观测信号被转换为多个频率特定的观测信号，每个频率特定的观测信号对应于多个频带之一。 然后，通过使用基于声源模型的频带的频率特定观察信号和表示音频信号中的每个频带的关系的混响模型来估计与每个频带相对应的混响滤波器， 观察信号和混响滤波器。 通过对频带应用频带的混响除复用滤波器来确定对应于每个频带的频率特定目标信号，并将所得到的频率特定目标信号进行积分。

公开(公告)号：US08467538B2

公开(公告)日：2013-06-18

申请号：US12919694

申请日：2009-02-27

申请人： Tomohiro Nakatani ,  Takuya Yoshioka ,  Keisuke Kinoshita ,  Masato Miyoshi

发明人： Tomohiro Nakatani ,  Takuya Yoshioka ,  Keisuke Kinoshita ,  Masato Miyoshi

IPC分类号： H04B3/20

CPC分类号： G10L2021/02082

摘要： A sound source model storage section stores a sound source model that represents an audio signal emitted from a sound source in the form of a probability density function. An observation signal, which is obtained by collecting the audio signal, is converted into a plurality of frequency-specific observation signals each corresponding to one of a plurality of frequency bands. Then, a dereverberation filter corresponding to each frequency band is estimated by using the frequency-specific observation signal for the frequency band on the basis of the sound source model and a reverberation model that represents a relationship for each frequency band among the audio signal, the observation signal and the dereverberation filter. A frequency-specific target signal corresponding to each frequency band is determined by applying the dereverberation filter for the frequency band to the frequency-specific observation signal for the frequency band, and the resulting frequency-specific target signals are integrated.

摘要翻译： 声源模型存储部存储以概率密度函数的形式表示从声源发出的音频信号的声源模型。 通过收集音频信号而获得的观测信号被转换为多个频率特定的观测信号，每个频率特定的观测信号对应于多个频带之一。 然后，通过使用基于声源模型的频带的频率特定观察信号和表示音频信号中的每个频带的关系的混响模型来估计与每个频带相对应的混响滤波器， 观察信号和混响滤波器。 通过对频带应用频带的混响除复用滤波器来确定对应于每个频带的频率特定目标信号，并将所得到的频率特定目标信号进行积分。

公开(公告)号：US08271277B2

公开(公告)日：2012-09-18

申请号：US12280101

申请日：2007-03-05

申请人： Keisuke Kinoshita ,  Tomohiro Nakatani ,  Masato Miyoshi

发明人： Keisuke Kinoshita ,  Tomohiro Nakatani ,  Masato Miyoshi

IPC分类号： G10L21/02 ,  H04B3/20

CPC分类号： H04N7/147 ,  G10L2021/02082 ,  H04R3/04 ,  H04S7/305

摘要： A model application unit calculates linear prediction coefficients of a multi-step linear prediction model by using discrete acoustic signals. Then, a late reverberation predictor calculates linear prediction values obtained by substituting the linear prediction coefficients and the discrete acoustic signals into linear prediction term of the multi-step linear prediction model, as predicted late reverberations. Next, a frequency domain converter converts the discrete acoustic signals to discrete acoustic signals in the frequency domain and also converts the predicted late reverberations to predicted late reverberations in the frequency domain. A late reverberation eliminator calculates relative values between the amplitude spectra of the discrete acoustic signals expressed in the frequency domain and the amplitude spectra of the predicted late reverberations expressed in the frequency domain, and provides the relative values as predicted amplitude spectra of a dereverberation signal.

摘要翻译： 模型应用单元通过使用离散声信号来计算多步线性预测模型的线性预测系数。 然后，后期混响预测器计算通过将线性预测系数和离散声信号代入多阶线性预测模型的线性预测项而获得的线性预测值，如预测的后期混响。 接下来，频域转换器将离散声学信号转换成频域中的离散声学信号，并将预测的后期混响转换为频域中预测的后期混响。 后期的混响消除器计算在频域中表示的离散声信号的幅度谱与在频域中表示的预测后期混响的幅度谱之间的相对值，并将相对值作为混响前信号的预测振幅谱提供。

公开(公告)号：US20090248403A1

公开(公告)日：2009-10-01

申请号：US12280101

申请日：2007-03-05

申请人： Keisuke Kinoshita ,  Tomohiro Nakatani ,  Masato Miyoshi

发明人： Keisuke Kinoshita ,  Tomohiro Nakatani ,  Masato Miyoshi

IPC分类号： G10L19/00

CPC分类号： H04N7/147 ,  G10L2021/02082 ,  H04R3/04 ,  H04S7/305

摘要： A model application unit calculates linear prediction coefficients of a multi-step linear prediction model by using discrete acoustic signals. Then, a late reverberation predictor calculates linear prediction values obtained by substituting the linear prediction coefficients and the discrete acoustic signals into linear prediction term of the multi-step linear prediction model, as predicted late reverberations. Next, a frequency domain converter converts the discrete acoustic signals to discrete acoustic signals in the frequency domain and also converts the predicted late reverberations to predicted late reverberations in the frequency domain. A late reverberation eliminator calculates relative values between the amplitude spectra of the discrete acoustic signals expressed in the frequency domain and the amplitude spectra of the predicted late reverberations expressed in the frequency domain, and provides the relative values as predicted amplitude spectra of a dereverberation signal.

摘要翻译： 模型应用单元通过使用离散声信号来计算多步线性预测模型的线性预测系数。 然后，后期混响预测器计算通过将线性预测系数和离散声信号代入多阶线性预测模型的线性预测项而获得的线性预测值，如预测的后期混响。 接下来，频域转换器将离散声学信号转换成频域中的离散声学信号，并将预测的后期混响转换为频域中预测的后期混响。 后期的混响消除器计算在频域中表示的离散声信号的幅度谱与在频域中表示的预测后期混响的幅度谱之间的相对值，并将相对值作为混响前信号的预测振幅谱提供。

公开(公告)号：US08848933B2

公开(公告)日：2014-09-30

申请号：US12920222

申请日：2009-03-05

申请人： Takuya Yoshioka ,  Tomohiro Nakatani ,  Masato Miyoshi

发明人： Takuya Yoshioka ,  Tomohiro Nakatani ,  Masato Miyoshi

IPC分类号： H04B3/20

CPC分类号： G10L2021/02082

摘要： The initial values of parameter estimates are set, including reverberation parameter estimates, which includes a regression coefficient used in a linear convolutional operation for calculating an estimated value of reverberation included in an observed signal, source parameter estimates, which includes estimated values of a linear prediction coefficient and a prediction residual power that identify the power spectrum of a source signal, and noise parameter estimates, which include noise power spectrum estimates. Then, the maximum likelihood estimation is used to alternately repeat processing for updating at least one of the reverberation parameter estimates and the noise parameter estimates and processing for updating the source parameter estimates until a predetermined termination condition is satisfied.

摘要翻译： 设置参数估计的初始值，包括混响参数估计，其包括在线性卷积运算中使用的回归系数，用于计算包括在观察信号中的混响的估计值，源参数估计，其包括线性预测的估计值 系数和识别源信号的功率谱的预测剩余功率，以及包括噪声功率谱估计的噪声参数估计。 然后，使用最大似然估计来交替地重复用于更新混响参数估计和噪声参数估计中的至少一个的处理，以及用于更新源参数估计的处理，直到满足预定的终止条件。

公开(公告)号：US20110044462A1

公开(公告)日：2011-02-24

申请号：US12920222

申请日：2009-03-05

申请人： Takuya Yoshioka ,  Tomohiro Nakatani ,  Masato Miyoshi

发明人： Takuya Yoshioka ,  Tomohiro Nakatani ,  Masato Miyoshi

IPC分类号： H04B3/20

CPC分类号： G10L2021/02082

摘要： The initial values of parameter estimates are set, including reverberation parameter estimates, which includes a regression coefficient used in a linear convolutional operation for calculating an estimated value of reverberation included in an observed signal, source parameter estimates, which includes estimated values of a linear prediction coefficient and a prediction residual power that identify the power spectrum of a source signal, and noise parameter estimates, which include noise power spectrum estimates. Then, the maximum likelihood estimation is used to alternately repeat processing for updating at least one of the reverberation parameter estimates and the noise parameter estimates and processing for updating the source parameter estimates until a predetermined termination condition is satisfied.

摘要翻译： 设置参数估计的初始值，包括混响参数估计，其包括在线性卷积运算中使用的回归系数，用于计算包括在观察信号中的混响的估计值，源参数估计，其包括线性预测的估计值 系数和识别源信号的功率谱的预测剩余功率，以及包括噪声功率谱估计的噪声参数估计。 然后，使用最大似然估计来交替地重复用于更新混响参数估计和噪声参数估计中的至少一个的处理，以及用于更新源参数估计的处理，直到满足预定的终止条件。

公开(公告)号：US08290170B2

公开(公告)日：2012-10-16

申请号：US12282762

申请日：2006-05-01

申请人： Tomohiro Nakatani ,  Biing-Hwang Juang

发明人： Tomohiro Nakatani ,  Biing-Hwang Juang

IPC分类号： H04B3/20

CPC分类号： G10L21/0232 ,  G10L2021/02082

摘要： Speech dereverberation is achieved by accepting an observed signal for initialization (1000) and performing likelihood maximization (2000) which includes Fourier Transforms (4000).

摘要翻译： 通过接受用于初始化的观测信号（1000）并执行包括傅里叶变换（4000）的似然最大化（2000））来实现语音去混响。

公开(公告)号：US20130168141A1

公开(公告)日：2013-07-04

申请号：US13820758

申请日：2012-01-24

申请人： Junichi Hozumi ,  Takumi Taura ,  Shin Okumura ,  Tomohiro Nakatani ,  Ryo Tomoida

发明人： Junichi Hozumi ,  Takumi Taura ,  Shin Okumura ,  Tomohiro Nakatani ,  Ryo Tomoida

IPC分类号： H01L21/762 ,  H05K1/03

CPC分类号： H01L21/76251 ,  B81C1/00095 ,  B81C1/00301 ,  C03C15/00 ,  G01P15/0802 ,  G01P15/125 ,  G01P2015/0834 ,  H01L21/486 ,  H01L23/057 ,  H01L23/10 ,  H01L23/15 ,  H01L2224/45144 ,  H01L2224/48091 ,  H01L2224/48137 ,  H01L2224/48247 ,  H01L2924/1461 ,  H05K1/0306 ,  H01L2924/00014 ,  H01L2924/00

摘要： A method for producing a substrate with through-electrode includes the steps of: forming recesses or through-holes in either one of a silicon substrate and a glass substrate; forming protrusions in the other substrate; laying the silicon substrate and glass substrate on each other so that the protrusions are inserted in the respective recesses or through-holes; and bonding the silicon substrate and the glass substrate to each other.

摘要翻译： 用于制造具有贯通电极的基板的方法包括以下步骤：在硅基板和玻璃基板中的任一个中形成凹部或通孔; 在另一基板上形成突起; 将硅基板和玻璃基板彼此铺设，使得突起插入相应的凹部或通孔中; 并将硅衬底和玻璃衬底彼此接合。

公开(公告)号：US20120173234A1

公开(公告)日：2012-07-05

申请号：US13384917

申请日：2010-07-15

申请人： Masakiyo Fujimoto ,  Tomohiro Nakatani

发明人： Masakiyo Fujimoto ,  Tomohiro Nakatani

IPC分类号： G10L15/20

CPC分类号： G10L15/20 ,  G10L15/14 ,  G10L21/02

摘要： The processing efficiency and estimation accuracy of a voice activity detection apparatus are improved. An acoustic signal analyzer receives a digital acoustic signal containing a speech signal and a noise signal, generates a non-speech GMM and a speech GMM adapted to a noise environment, by using a silence GMM and a clean-speech GMM in each frame of the digital acoustic signal, and calculates the output probabilities of dominant Gaussian distributions of the GMMs. A speech state probability to non-speech state probability ratio calculator calculates a speech state probability to non-speech state probability ratio based on a state transition model of a speech state and a non-speech state, by using the output probabilities; and a voice activity detection unit judges, from the speech state probability to non-speech state probability ratio, whether the acoustic signal in the frame is in the speech state or in the non-speech state and outputs only the acoustic signal in the speech state.

摘要翻译： 提高语音活动检测装置的处理效率和估计精度。 声信号分析仪接收包含语音信号和噪声信号的数字声信号，通过使用静音GMM和干扰语音GMM在每个帧中产生适用于噪声环境的非语音GMM和语音GMM 数字声信号，并计算GMM的显性高斯分布的输出概率。 通过使用输出概率，基于语音状态和非语音状态的状态转换模型，非语音状态概率比计算器的语音状态概率将语音状态概率计算为非语音状态概率比; 并且语音活动检测单元从语音状态概率到非语音状态概率比判断帧中的声音信号是处于语音状态还是处于非语音状态，并且仅输出语音状态下的声音信号 。

公开(公告)号：US09208780B2

公开(公告)日：2015-12-08

申请号：US13384917

申请日：2010-07-15

申请人： Masakiyo Fujimoto ,  Tomohiro Nakatani

发明人： Masakiyo Fujimoto ,  Tomohiro Nakatani

IPC分类号： G10L15/20 ,  G10L15/14 ,  G10L21/02

CPC分类号： G10L15/20 ,  G10L15/14 ,  G10L21/02

摘要： The processing efficiency and estimation accuracy of a voice activity detection apparatus are improved. An acoustic signal analyzer receives a digital acoustic signal containing a speech signal and a noise signal, generates a non-speech GMM and a speech GMM adapted to a noise environment, by using a silence GMM and a clean-speech GMM in each frame of the digital acoustic signal, and calculates the output probabilities of dominant Gaussian distributions of the GMMs. A speech state probability to non-speech state probability ratio calculator calculates a speech state probability to non-speech state probability ratio based on a state transition model of a speech state and a non-speech state, by using the output probabilities; and a voice activity detection unit judges, from the speech state probability to non-speech state probability ratio, whether the acoustic signal in the frame is in the speech state or in the non-speech state and outputs only the acoustic signal in the speech state.

摘要翻译： 提高语音活动检测装置的处理效率和估计精度。 声信号分析仪接收包含语音信号和噪声信号的数字声信号，通过使用静音GMM和干扰语音GMM在每个帧中产生适用于噪声环境的非语音GMM和语音GMM 数字声信号，并计算GMM的显性高斯分布的输出概率。 通过使用输出概率，基于语音状态和非语音状态的状态转换模型，非语音状态概率比计算器的语音状态概率将语音状态概率计算为非语音状态概率比; 并且语音活动检测单元从语音状态概率到非语音状态概率比判断帧中的声音信号是处于语音状态还是处于非语音状态，并且仅输出语音状态下的声音信号 。