公开(公告)号：US09407990B2

公开(公告)日：2016-08-02

申请号：US12892078

申请日：2010-09-28

申请人： Kyu-Hong Kim ,  So-Young Jeong ,  Kwang-Cheol Oh ,  Jae-Hoon Jeong

发明人： Kyu-Hong Kim ,  So-Young Jeong ,  Kwang-Cheol Oh ,  Jae-Hoon Jeong

IPC分类号： H04R3/00 ,  H04R29/00

CPC分类号： H04R3/005 ,  H04R29/004 ,  H04R2201/401

摘要： An apparatus and method for calibrating gain difference between microphones included in a microphone array are provided. In the gain calibrating apparatus, weights for each frequency component of the acoustic signals, which have been converted into the signals in the frequency domain are calculated. The weights are used to calibrate the acoustic signals such that the plurality of acoustic signals each have the same amplitude while the acoustic signals maintain their individual phase. The amplitudes of the acoustic signals are calibrated by use of the calculated weights. The gain calibrating apparatus calibrates gain in real time while calculating weights for frequency components of the frame of acoustic signals in real time.

摘要翻译： 提供了一种用于校准包括在麦克风阵列中的麦克风之间的增益差的装置和方法。 在增益校准装置中，计算已被转换成频域信号的声信号的每个频率分量的权重。 权重用于校准声信号，使得多个声信号各自具有相同的幅度，同时声信号保持其各自的相位。 通过使用计算的权重校正声信号的振幅。 增益校准装置实时校准增益，同时实时计算声信号帧的频率分量的权重。

公开(公告)号：US07818169B2

公开(公告)日：2010-10-19

申请号：US11649161

申请日：2007-01-04

申请人： Kwang Cheol Oh ,  Jae-Hoon Jeong ,  So-Young Jeong

发明人： Kwang Cheol Oh ,  Jae-Hoon Jeong ,  So-Young Jeong

IPC分类号： G10L15/00

CPC分类号： G10L15/02 ,  G10L25/15 ,  G10L25/90

摘要： A formant frequency estimation method which is important information in speech recognition by accelerating a spectrum using a pitch frequency, and an apparatus using the method is provided. That is, the formant frequency estimation method includes preprocessing an input speech signal and generating a spectrum by a fast Fourier transforming the preprocessed input speech signal; smoothing the generated spectrum; accelerating the smoothed spectrum; and determining a formant frequency on the basis of the accelerated spectrum.

摘要翻译： 提供了通过使用音调频率加速频谱的语音识别中的重要信息的共振峰频率估计方法，以及使用该方法的装置。 也就是说，共振峰频率估计方法包括对输入语音信号进行预处理，并通过对预处理的输入语音信号进行快速傅立叶变换来产生频谱; 平滑生成的光谱; 加速平滑光谱; 以及基于加速频谱确定共振峰频率。

公开(公告)号：US20070192088A1

公开(公告)日：2007-08-16

申请号：US11649161

申请日：2007-01-04

申请人： Kwang Cheol Oh ,  Jae-Hoon Jeong ,  So-Young Jeong

发明人： Kwang Cheol Oh ,  Jae-Hoon Jeong ,  So-Young Jeong

IPC分类号： G10L19/06

CPC分类号： G10L15/02 ,  G10L25/15 ,  G10L25/90

摘要： A formant frequency estimation method which is important information in speech recognition by accelerating a spectrum using a pitch frequency, and an apparatus using the method is provided. That is, the formant frequency estimation method includes preprocessing an input speech signal and generating a spectrum by a fast Fourier transforming the preprocessed input speech signal; smoothing the generated spectrum; accelerating the smoothed spectrum; and determining a formant frequency on the basis of the accelerated spectrum.

摘要翻译： 提供了通过使用音调频率加速频谱的语音识别中的重要信息的共振峰频率估计方法，以及使用该方法的装置。 也就是说，共振峰频率估计方法包括对输入语音信号进行预处理，并通过对预处理的输入语音信号进行快速傅立叶变换来产生频谱; 平滑生成的光谱; 加速平滑光谱; 以及基于加速频谱确定共振峰频率。

公开(公告)号：US08665305B2

公开(公告)日：2014-03-04

申请号：US13181568

申请日：2011-07-13

申请人： Jae-Hoon Jeong ,  Jeong-Su Kim ,  Kwang-Cheol Oh ,  So-Young Jeong

发明人： Jae-Hoon Jeong ,  Jeong-Su Kim ,  Kwang-Cheol Oh ,  So-Young Jeong

IPC分类号： H04N7/14

CPC分类号： H04N7/14 ,  H04N7/147 ,  H04R3/005 ,  H04R2410/01 ,  H04R2430/03 ,  H04R2499/11

摘要： A microphone array apparatus is provided. The microphone array apparatus includes a plurality of first type microphones which are disposed to be hidden in a direction of a target source and surrounded by a cover that passes acoustic signals therethrough, and a plurality of second type microphones which are disposed at both sides of the plurality of first type microphones.

摘要翻译： 提供麦克风阵列装置。 麦克风阵列装置包括多个第一类型麦克风，其被设置为隐藏在目标源的方向上并且被传导声信号的盖包围，并且多个第二类型麦克风被布置在 多个第一类麦克风。

公开(公告)号：US20120179458A1

公开(公告)日：2012-07-12

申请号：US13286369

申请日：2011-11-01

申请人： Kwang-Cheol Oh ,  Jeong-Su Kim ,  Jae-Hoon Jeong ,  So-Young Jeong

发明人： Kwang-Cheol Oh ,  Jeong-Su Kim ,  Jae-Hoon Jeong ,  So-Young Jeong

IPC分类号： G10L21/02

CPC分类号： G10L21/0208 ,  G10L25/84 ,  G10L2021/02166

摘要： Provided are an apparatus and method for estimating noise that changes with time. The apparatus may calculate a speech absence probability that indicates the possibility of the absence of speech in each frequency component of an input acoustic signal, may discriminate between a speech-dominant region and a noise region from the acoustic signals based on the speech absence probability, and may estimate noise according to the discrimination result.

摘要翻译： 提供了一种用于估计随时间变化的噪声的装置和方法。 该装置可以计算表示在输入声信号的每个频率分量中不存在语音的可能性的语音缺失概率，可以基于语音不存在概率从声信号中区分语音主导区域和噪声区域， 并且可以根据判别结果估计噪声。

公开(公告)号：US20110075859A1

公开(公告)日：2011-03-31

申请号：US12892078

申请日：2010-09-28

申请人： Kyu-Hong KIM ,  So-Young Jeong ,  Kwang-Cheol Oh ,  Jae-Hoon Jeong

发明人： Kyu-Hong KIM ,  So-Young Jeong ,  Kwang-Cheol Oh ,  Jae-Hoon Jeong

IPC分类号： H04R3/00

CPC分类号： H04R3/005 ,  H04R29/004 ,  H04R2201/401

摘要： An apparatus and method for calibrating gain difference between microphones included in a microphone array are provided. In the gain calibrating apparatus, weights for each frequency component of the acoustic signals, which have been converted into the signals in the frequency domain are calculated. The weights are used to calibrate the acoustic signals such that the plurality of acoustic signals each have the same amplitude while the acoustic signals maintain their individual phase. The amplitudes of the acoustic signals are calibrated by use of the calculated weights. The gain calibrating apparatus calibrates gain in real time while calculating weights for frequency components of the frame of acoustic signals in real time.

摘要翻译： 提供了一种用于校准包括在麦克风阵列中的麦克风之间的增益差的装置和方法。 在增益校准装置中，计算已被转换成频域信号的声信号的每个频率分量的权重。 权重用于校准声信号，使得多个声信号各自具有相同的幅度，同时声信号保持其各自的相位。 通过使用计算的权重校正声信号的振幅。 增益校准装置实时校准增益，同时实时计算声信号帧的频率分量的权重。

公开(公告)号：US08965002B2

公开(公告)日：2015-02-24

申请号：US13114746

申请日：2011-05-24

申请人： Kwang-Cheol Oh ,  Jeong-Su Kim ,  Jae-Hoon Jeong ,  So-Young Jeong

发明人： Kwang-Cheol Oh ,  Jeong-Su Kim ,  Jae-Hoon Jeong ,  So-Young Jeong

IPC分类号： H04R3/00 ,  G10L21/0208 ,  G10L21/0216

CPC分类号： G10L21/0208 ,  G10L2021/02166

摘要： An audio quality enhancing apparatus and method is provided in which a microphone array has a non-uniform configuration and thus a beam pattern of a desired direction is obtained in a wide range of frequencies including higher frequency bands and lower frequency bands even when the microphone array is relatively small. The audio quality enhancing apparatus includes at least three microphones which are disposed in a non-uniform configuration, a frequency conversion unit configured to transform acoustic signals input from the at least three microphones to acoustic signals of frequency domain; a band division and merging unit configured to divide frequencies of the transformed acoustic signals into bands based on intervals between the at least three microphones and to merge the acoustic signals in the frequency domain into signals of two channels based on the divided frequency bands; and a two channel beamforming unit configured to reduce noise of signals including input from a direction other than the direction of a target sound by performing beamforming on the signals of the two channels and to output the noise-reduced signals.

摘要翻译： 提供了一种音频质量增强装置和方法，其中麦克风阵列具有不均匀的配置，因此即使在麦克风阵列中也可以在包括较高频带和较低频带的宽频率范围内获得期望方向的波束图案 比较小 音频质量提高装置包括以非均匀配置布置的至少三个麦克风，被配置为将从至少三个麦克风输入的声信号变换为频域的声信号的频率转换单元; 频带划分和合并单元，被配置为基于所述至少三个麦克风之间的间隔将所述经变换的声信号的频率划分成频带，并且基于所划分的频带将频域中的声信号合并成两个信道的信号; 以及双通道波束形成单元，被配置为通过对所述两个通道的信号执行波束形成并且输出所述降噪信号，来减少包括来自目标声音的方向的方向的输入的信号的噪声。

公开(公告)号：US07792669B2

公开(公告)日：2010-09-07

申请号：US11657654

申请日：2007-01-25

申请人： Kwang Cheol Oh ,  Jae-Hoon Jeong

发明人： Kwang Cheol Oh ,  Jae-Hoon Jeong

IPC分类号： G10L11/06

CPC分类号： G10L25/93 ,  G10L25/06

摘要： A method and apparatus of estimating a voicing for speech recognition by using local spectral information. The voicing estimation method for speech recognition includes performing a Fourier transform on input voice signals after performing pre-processing on the input voice signals. The method further includes detecting peaks in the input voice signals after smoothing the input voice signals. The method also includes computing every frequency bound associated with the detected peaks, and determining a class of a voicing according to each computed frequency bound.

摘要翻译： 通过使用局部光谱信息来估计用于语音识别的语音的方法和装置。 用于语音识别的语音预测方法包括在对输入的语音信号进行预处理之后对输入语音信号进行傅立叶变换。 该方法还包括在平滑输入的语音信号之后检测输入语音信号中的峰值。 该方法还包括计算与检测到的峰值相关联的每个频率范围，以及根据每个计算出的频率边界来确定语音类别。

公开(公告)号：US20070185712A1

公开(公告)日：2007-08-09

申请号：US11477628

申请日：2006-06-30

申请人： Jae-Hoon Jeong ,  Kwang Cheol Oh

发明人： Jae-Hoon Jeong ,  Kwang Cheol Oh

IPC分类号： G10L15/00

CPC分类号： G10L15/01

摘要： A method of measuring confidence of speech recognition in a speech recognizer compares a phase change point with a phoneme string change point and uses a difference between the phase change point and the phoneme string change point and a likelihood ratio, and an apparatus using the method is provided. That is, the method of the present invention includes detecting a phase change point of a speech signal; detecting a phoneme string change point according to a result of speech recognition; calculating confidence of the speech recognition by using a difference between the detected phase change point and phoneme string change point. According to the present invention, a performance of measuring confidence may become improved by simultaneously taking not only a likelihood ratio, but also taking a comparison result of a phase change point with a phoneme string change point into consideration.

摘要翻译： 一种测量语音识别器中语音识别的置信度的方法将相变点与音素串变化点进行比较，并使用相位变化点和音素串变化点之间的差异以及似然比，并且使用该方法的装置 提供。 也就是说，本发明的方法包括检测语音信号的相变点; 根据语音识别的结果检测音素串变化点; 通过使用检测到的相变点和音素串变化点之间的差来计算语音识别的置信度。 根据本发明，通过不仅考虑似然比，而且考虑到具有音素串变化点的相变点的比较结果，可以提高测量置信度的性能。

公开(公告)号：US20070185709A1

公开(公告)日：2007-08-09

申请号：US11657654

申请日：2007-01-25

申请人： Kwang Cheol Oh ,  Jae-Hoon Jeong

发明人： Kwang Cheol Oh ,  Jae-Hoon Jeong

IPC分类号： G10L11/06

CPC分类号： G10L25/93 ,  G10L25/06

摘要： A method and apparatus of estimating a voicing for speech recognition by using local spectral information. The voicing estimation method for speech recognition includes performing a Fourier transform on input voice signals after performing pre-processing on the input voice signals. The method further includes detecting peaks in the input voice signals after smoothing the input voice signals. The method also includes computing every frequency bound associated with the detected peaks, and determining a class of a voicing according to each computed frequency bound.

摘要翻译： 通过使用局部光谱信息来估计用于语音识别的语音的方法和装置。 用于语音识别的语音预测方法包括在对输入的语音信号进行预处理之后对输入语音信号进行傅立叶变换。 该方法还包括在平滑输入的语音信号之后检测输入语音信号中的峰值。 该方法还包括计算与检测到的峰值相关联的每个频率范围，以及根据每个计算出的频率边界来确定语音类别。