-
1.发明申请Apparatus and method for speech recognition using a plurality of confidence score estimation algorithms 有权使用多个置信分数估计算法进行语音识别的装置和方法公开(公告)号：US20070136058A1
公开(公告)日：2007-06-14
申请号：US11517369
申请日：2006-09-08
申请人： Jae-hoon Jeong , Sang-bae Jeong , Jeong-su Kim , Nam-hoon Kim
发明人： Jae-hoon Jeong , Sang-bae Jeong , Jeong-su Kim , Nam-hoon Kim
IPC分类号： G10L15/00
CPC分类号： G10L15/08 , G10L2015/088
摘要： An apparatus for speech recognition includes: a first confidence score calculator calculating a first confidence score using a ratio between a likelihood of a keyword model for feature vectors per frame of a speech signal and a likelihood of a Filler model for the feature vectors; a second confidence score calculator calculating a second confidence score by comparing a Gaussian distribution trace of the keyword model per frame of the speech signal with a Gaussian distribution trace sample of a stored corresponding keyword of the keyword model; and a determination module determining a confidence of a result using the keyword model in accordance with a position determined by the first and second confidence scores on a confidence coordinate system.
摘要翻译： 一种用于语音识别的装置包括：第一置信度分数计算器，使用针对每个语音信号的每个特征向量的关键字模型的似然率与特征向量的填充模型的似然率之间的比率来计算第一置信度分数; 第二置信度计算器通过将所述语音信号的每帧的关键字模型的高斯分布轨迹与所述关键字模型的存储的对应关键字的高斯分布轨迹样本进行比较来计算第二置信度分数; 以及确定模块，其根据由置信坐标系上的第一和第二置信度得分确定的位置，使用关键字模型确定结果的置信度。
-
2.发明授权Apparatus and method for speech recognition using a plurality of confidence score estimation algorithms 有权使用多个置信分数估计算法进行语音识别的装置和方法公开(公告)号：US08543399B2
公开(公告)日：2013-09-24
申请号：US11517369
申请日：2006-09-08
申请人： Jae-hoon Jeong , Sang-bae Jeong , Jeong-su Kim , Nam-hoon Kim
发明人： Jae-hoon Jeong , Sang-bae Jeong , Jeong-su Kim , Nam-hoon Kim
IPC分类号： G10L15/00
CPC分类号： G10L15/08 , G10L2015/088
摘要： An apparatus for speech recognition includes: a first confidence score calculator calculating a first confidence score using a ratio between a likelihood of a keyword model for feature vectors per frame of a speech signal and a likelihood of a Filler model for the feature vectors; a second confidence score calculator calculating a second confidence score by comparing a Gaussian distribution trace of the keyword model per frame of the speech signal with a Gaussian distribution trace sample of a stored corresponding keyword of the keyword model; and a determination module determining a confidence of a result using the keyword model in accordance with a position determined by the first and second confidence scores on a confidence coordinate system.
摘要翻译： 一种用于语音识别的装置包括：第一置信度分数计算器，使用针对每个语音信号的每个特征向量的关键字模型的似然率与特征向量的填充模型的似然率之间的比率来计算第一置信度分数; 第二置信度计算器通过将所述语音信号的每帧的关键字模型的高斯分布轨迹与所述关键字模型的存储的对应关键字的高斯分布轨迹样本进行比较来计算第二置信度分数; 以及确定模块，其根据由置信坐标系上的第一和第二置信度得分确定的位置，使用关键字模型确定结果的置信度。
-
3.发明授权Method and apparatus for recognizing speech by measuring confidence levels of respective frames 有权通过测量各帧的置信水平来识别语音的方法和装置公开(公告)号：US08271283B2
公开(公告)日：2012-09-18
申请号：US11355082
申请日：2006-02-16
申请人： Jae-hoon Jeong , Sang-bae Jeong , Jeong-su Kim , In-Jeong Choi
发明人： Jae-hoon Jeong , Sang-bae Jeong , Jeong-su Kim , In-Jeong Choi
CPC分类号： G10L15/08 , G10L15/142
摘要： Disclosed herein is a method and apparatus to recognize speech by measuring the confidence levels of respective frames. The method includes the operations of obtaining frequency features of a received speech signal for the respective frames having a predetermined length, calculating a keyword model-based likelihood and a filler model-based likelihood for each of the frame, calculating a confidence score based on the two types of likelihoods, and deciding whether the received speech signal corresponds to a keyword or a non-keyword based on the confidence scores. Also, the method includes the operation of transforming the confidence scores by applying transform functions of clusters, which include the confidence scores or are close to the confidence scores, to the confidence scores.
摘要翻译： 本文公开了一种通过测量各个帧的置信水平来识别语音的方法和装置。 该方法包括获得具有预定长度的各个帧的接收到的语音信号的频率特征的操作，计算每个帧的基于关键词模型的可能性和基于填充模型的可能性，基于 两种类型的可能性，并且基于置信度分数来决定接收到的语音信号是否对应于关键字或非关键字。 此外，该方法包括通过将包括置信分数或接近置信度得分的聚类的变换函数应用到置信度得分来变换置信度分数的操作。
-
4.发明申请Method and apparatus for recognizing speech by measuring confidence levels of respective frames 有权通过测量各帧的置信水平来识别语音的方法和装置公开(公告)号：US20060190259A1
公开(公告)日：2006-08-24
申请号：US11355082
申请日：2006-02-16
申请人： Jae-hoon Jeong , Sang-bae Jeong , Jeong-su Kim , In-Jeong Choi
发明人： Jae-hoon Jeong , Sang-bae Jeong , Jeong-su Kim , In-Jeong Choi
IPC分类号： G10L15/14
CPC分类号： G10L15/08 , G10L15/142
摘要： Disclosed herein is a method and apparatus to recognize speech by measuring the confidence levels of respective frames. The method includes the operations of obtaining frequency features of a received speech signal for the respective frames having a predetermined length, calculating a keyword model-based likelihood and a filler model-based likelihood for each of the frame, calculating a confidence score based on the two types of likelihoods, and deciding whether the received speech signal corresponds to a keyword or a non-keyword based on the confidence scores. Also, the method includes the operation of transforming the confidence scores by applying transform functions of clusters, which include the confidence scores or are close to the confidence scores, to the confidence scores.
摘要翻译： 本文公开了一种通过测量各个帧的置信水平来识别语音的方法和装置。 该方法包括获得具有预定长度的各个帧的接收到的语音信号的频率特征的操作，计算每个帧的基于关键词模型的可能性和基于填充模型的可能性，基于 两种类型的可能性，并且基于置信度分数来决定接收到的语音信号是否对应于关键字或非关键字。 此外，该方法包括通过将包括置信分数或接近置信度得分的聚类的变换函数应用到置信度得分来变换置信度分数的操作。
-
公开(公告)号：US20080147391A1
公开(公告)日：2008-06-19
申请号：US11896450
申请日：2007-08-31
申请人： So-young Jeong , Kwang-cheol Oh , Jae-hoon Jeong , Jeong-su Kim
发明人： So-young Jeong , Kwang-cheol Oh , Jae-hoon Jeong , Jeong-su Kim
IPC分类号： G10L15/16
摘要： Provided is a method and apparatus for transforming a speech feature vector. The method includes extracting a feature vector required for speech recognition from a speech signal and transforming the extracted feature vector using an auto-associative neural network (AANN).
摘要翻译： 提供了一种用于变换语音特征向量的方法和装置。 该方法包括从语音信号中提取语音识别所需的特征向量，并使用自相关神经网络（AANN）对所提取的特征向量进行变换。
-
公开(公告)号：US20080208577A1
公开(公告)日：2008-08-28
申请号：US11889665
申请日：2007-08-15
申请人： So-young Jeong , Kwang-cheol Oh , Jae-hoon Jeong , Jeong-su Kim
发明人： So-young Jeong , Kwang-cheol Oh , Jae-hoon Jeong , Jeong-su Kim
IPC分类号： G10L15/00
摘要： Provided are a multi-stage speech recognition apparatus and method. The multi-stage speech recognition apparatus includes a first speech recognition unit performing initial speech recognition on a feature vector, which is extracted from an input speech signal, and generating a plurality of candidate words; and a second speech recognition unit rescoring the candidate words, which are provided by the first speech recognition unit, using a temporal posterior feature vector extracted from the speech signal.
摘要翻译： 提供了一种多级语音识别装置和方法。 多级语音识别装置包括：第一语音识别单元，对从输入语音信号提取的特征向量进行初始语音识别，生成多个候选词; 以及第二语音识别单元，使用从所述语音信号提取的时间后向特征向量，对由所述第一语音识别单元提供的候选词进行重新排序。
-
7.发明授权Audio signal processing method, audio apparatus therefor, and electronic apparatus therefor 有权音频信号处理方法，音频装置及其电子设备公开(公告)号：US09047862B2
公开(公告)日：2015-06-02
申请号：US13483571
申请日：2012-05-30
申请人： Jae-hoon Jeong , So-young Jeong , Jeong-su Kim , Jung-eun Park , Woo-jung Lee
发明人： Jae-hoon Jeong , So-young Jeong , Jeong-su Kim , Jung-eun Park , Woo-jung Lee
IPC分类号： G10L21/02 , H04B3/20 , G10L19/008 , G10L15/20 , G10L21/0208
CPC分类号： G10L19/008 , G10L15/20 , G10L2021/02082
摘要： An audio apparatus including a decorrelator for generating decorrelated signals by applying a phase shifting value adjusted based on a correlation difference between audio signals included in a multi-channel signal to the audio signals; and a speaker set including at least two speakers for outputting acoustic signals corresponding to the decorrelated signals.
摘要翻译： 一种音频装置，包括去相关器，用于通过将基于多声道信号中包括的音频信号之间的相关差调整的相移值应用于音频信号来产生去相关信号; 以及包括至少两个扬声器的扬声器组，用于输出对应于解相关信号的声信号。
-
8.发明授权Method and apparatus of transforming speech feature vectors using an auto-associative neural network 有权使用自动关联神经网络来转换语音特征向量的方法和装置公开(公告)号：US08838446B2
公开(公告)日：2014-09-16
申请号：US11896450
申请日：2007-08-31
申请人： So-young Jeong , Kwang-cheol Oh , Jae-hoon Jeong , Jeong-su Kim
发明人： So-young Jeong , Kwang-cheol Oh , Jae-hoon Jeong , Jeong-su Kim
摘要： Provided is a method and apparatus for transforming a speech feature vector. The method includes extracting a feature vector required for speech recognition from a speech signal and transforming the extracted feature vector using an auto-associative neural network (AANN).
摘要翻译： 提供了一种用于变换语音特征向量的方法和装置。 该方法包括从语音信号中提取语音识别所需的特征向量，并使用自相关神经网络（AANN）对所提取的特征向量进行变换。
-
公开(公告)号：US08762142B2
公开(公告)日：2014-06-24
申请号：US11889665
申请日：2007-08-15
申请人： So-young Jeong , Kwang-cheol Oh , Jae-hoon Jeong , Jeong-su Kim
发明人： So-young Jeong , Kwang-cheol Oh , Jae-hoon Jeong , Jeong-su Kim
摘要： Provided are a multi-stage speech recognition apparatus and method. The multi-stage speech recognition apparatus includes a first speech recognition unit performing initial speech recognition on a feature vector, which is extracted from an input speech signal, and generating a plurality of candidate words; and a second speech recognition unit rescoring the candidate words, which are provided by the first speech recognition unit, using a temporal posterior feature vector extracted from the speech signal.
摘要翻译： 提供了一种多级语音识别装置和方法。 多级语音识别装置包括：第一语音识别单元，对从输入语音信号提取的特征向量进行初始语音识别，生成多个候选词; 以及第二语音识别单元，使用从所述语音信号提取的时间后向特征向量，对由所述第一语音识别单元提供的候选词进行重新排序。
-
10.发明授权Sound source signal filtering method based on calculated distances between microphone and sound source 有权基于麦克风和声源之间计算距离的声源信号滤波方法公开(公告)号：US08385562B2
公开(公告)日：2013-02-26
申请号：US12149521
申请日：2008-05-02
申请人： Jae-hoon Jeong , Kwang-cheol Oh , Kyu-hong Kim , So-young Jeong
发明人： Jae-hoon Jeong , Kwang-cheol Oh , Kyu-hong Kim , So-young Jeong
IPC分类号： H04R3/00
摘要： Provided is a sound source signal filtering method and apparatus. The sound source signal filtering method includes: generating two or more microphone output signals by combining sound source signals input through a plurality of microphones; calculating distances between the microphones and a sound source from which the sound source signals are emitted by using distance relationships according to frequencies of the sound source signals extracted from the generated microphone output signals; and filtering the sound source signals to obtain one or more sound source signals corresponding to a predetermined distance by using the calculated distances. Accordingly, it is possible to obtain only sound source signals emitted from a sound source at a particular distance from the microphone array among a plurality of sound source signals input through the microphone array.
摘要翻译： 提供了一种声源信号滤波方法和装置。 声源信号滤波方法包括：通过组合通过多个麦克风输入的声源信号来产生两个或更多麦克风输出信号; 根据从产生的麦克风输出信号中提取的声源信号的频率，通过使用距离关系计算麦克风和声源之间的距离; 并且通过使用所计算的距离来对声源信号进行滤波以获得对应于预定距离的一个或多个声源信号。 因此，可以从通过麦克风阵列输入的多个声源信号中仅获得与麦克风阵列特定距离处的声源发出的声源信号。
-
-
-
-
-
-
-
-
-
搜索结果
30 条记录 (耗时 0.025 秒)
