公开(公告)号：US20060031071A1

公开(公告)日：2006-02-09

申请号：US10909997

申请日：2004-08-03

申请人： Gustavo Abrego ,  Atsuo Hiroe ,  Eugene Koontz

发明人： Gustavo Abrego ,  Atsuo Hiroe ,  Eugene Koontz

IPC分类号： G10L15/14

CPC分类号： G10L15/193

摘要： A system and method for automatically implementing a finite state automaton for speech recognition includes a finite state automaton generator that analyzes one or more input text sequences and automatically creates a node table and a link table to define the finite state automaton. The node table includes N-tuples from the input text sequences. Each N-tuple includes a current word and a corresponding history of one or more prior words from the input text sequences. The node table also includes unique node identifiers that each correspond to a different respective one of the current words. The link table includes specific links between successive words from the input text sequences. The links identified in the link table are defined by utilizing start node identifiers and end node identifiers from the unique node identifiers of the node table.

摘要翻译： 用于自动实现用于语音识别的有限状态自动机的系统和方法包括分析一个或多个输入文本序列并自动创建节点表和链接表以定义有限状态自动机的有限状态自动机生成器。 节点表包括来自输入文本序列的N元组。 每个N元组包括来自输入文本序列的当前单词和一个或多个先前单词的对应历史。 节点表还包括每个对应于当前单词中不同的相应一个的唯一节点标识符。 链接表包括来自输入文本序列的连续词之间的特定链接。 通过从节点表的唯一节点标识符中使用起始节点标识符和结束节点标识符来定义链接表中标识的链接。

公开(公告)号：US08358563B2

公开(公告)日：2013-01-22

申请号：US12480194

申请日：2009-06-08

申请人： Atsuo Hiroe

发明人： Atsuo Hiroe

IPC分类号： G01S3/80

CPC分类号： G01S3/8006 ,  G01S3/801

摘要： A signal processing apparatus includes: a learning processing unit that finds a separating matrix for separating mixed signals in which outputs from a plurality of sound sources are mixed, by a learning process that applies ICA (Independent Component Analysis) to observed signals including the mixed signals; a separation processing unit that applies the separating matrix to the observed signals to separate the mixed signals and generate separated signals corresponding to each of the sound sources; and a sound source direction estimating unit that computes a sound source direction of each of the generated separated signals. The sound source direction estimating unit calculates cross-covariance matrices between the observed signals and the separated signals in corresponding time segments in time-frequency domain, computes phase differences between elements of the cross-covariance matrices, and computes a sound source direction corresponding to each of the separated signals by applying the computed phase differences.

摘要翻译： 信号处理装置包括：学习处理单元，其通过对包括混合信号的观测信号应用ICA（独立分量分析）的学习处理，找到用于分离来自多个声源的输出的混合信号的分离矩阵 ; 分离处理单元，其将所述分离矩阵应用于所述观测信号，以分离所述混合信号并产生与每个所述声源相对应的分离信号; 以及声源方向估计单元，其计算每个所产生的分离信号的声源方向。 声源方向估计单元在时频域中计算观测信号与相应时间段中的分离信号之间的互协方差矩阵，计算交叉协方差矩阵的元素之间的相位差，并计算对应于每个 通过应用计算出的相位差来分离信号。

公开(公告)号：US07912696B1

公开(公告)日：2011-03-22

申请号：US09530200

申请日：1999-08-31

申请人： Yasuharu Asano ,  Atsuo Hiroe ,  Masato Shimakawa ,  Tetsuya Kagami ,  Erika Kobayashi

发明人： Yasuharu Asano ,  Atsuo Hiroe ,  Masato Shimakawa ,  Tetsuya Kagami ,  Erika Kobayashi

IPC分类号： G06F17/28

CPC分类号： G06F17/2836 ,  G06F17/289

摘要： A natural language processing apparatus includes an input section for inputting natural language, a representation converting section for converting representation of the natural language, a display section for displaying, for confirmation, sentence converted at the representation converting section, a machine translation section for carrying out machine translation of the confirmed sentence, and a control section for controlling these respective sections, thus to provide natural language processing in which confirmation operation of user is reduced.

摘要翻译： 自然语言处理装置包括用于输入自然语言的输入部分，用于转换自然语言的表示的表示转换部分，用于在表示转换部分转换的用于确认语句的转换的显示部分，用于执行自动语言处理的机器翻译部分 确定句子的机器翻译，以及用于控制这些各个部分的控制部分，从而提供减少用户确认操作的自然语言处理。

公开(公告)号：US20100278357A1

公开(公告)日：2010-11-04

申请号：US12661635

申请日：2010-03-22

申请人： Atsuo Hiroe

发明人： Atsuo Hiroe

IPC分类号： H04R3/00

CPC分类号： G10L21/0272

摘要： A signal processing apparatus includes a source separation module for producing respective separation signals corresponding to a plurality of sound sources by applying an ICA (Independent Component Analysis) to observation signals produced based on mixture signals from the sound sources, which are taken by source separation microphones, to thereby execute a separation process of the mixture signals, and a signal projection-back module for receiving observation signals of projection-back target microphones and the separation signals produced by the source separation module, and for producing projection-back signals as respective separation signals corresponding to the sound sources, which are taken by the projection-back target microphones. The signal projection-back module produces the projection-back signals by receiving the observation signals of the projection-back target microphones which differ from the source separation microphones.

摘要翻译： 信号处理装置包括：源分离模块，用于通过对由源分离麦克风采取的来自声源的混合信号产生的观测信号应用ICA（独立分量分析）来产生对应于多个声源的各个分离信号 ，从而执行混合信号的分离处理，以及用于接收投射返回目标麦克风的观测信号和由源分离模块产生的分离信号的信号投射返回模块，并且用于产生作为各自分离的投射返回信号 对应于声源的信号，其由投射回目标麦克风拍摄。 信号投射返回模块通过接收与源分离麦克风不同的投射返回目标麦克风的观测信号来产生投射返回信号。

公开(公告)号：US07809146B2

公开(公告)日：2010-10-05

申请号：US11421619

申请日：2006-06-01

申请人： Atsuo Hiroe ,  Keiichi Yamada

发明人： Atsuo Hiroe ,  Keiichi Yamada

IPC分类号： H04B15/00 ,  G10L19/02

CPC分类号： G10L21/0272

摘要： Problems of permutation can be solved with high accuracy without utilizing knowledge about original signals or information concerning positions of microphones and the like when each one of plural signals mixed in an audio signal is separated using independent component analysis. A short-time Fourier transformation section generates spectrograms of observation signals from observation signals in time domain. A signal separation section separates the spectrograms of the observation signals into spectrograms of respective signals, to generate spectrograms of separate signals. A permutation problem solution section calculates a scale corresponding to the degree of permutation, e.g., a Kullback-Leiblar information amount calculated by use of a multidimensional probability density function or multidimensional kurtosis, from substantial whole of the spectrograms of the separate signals. Based on the scale, signals at each of frequencies bin of the spectrograms of the separate signals are exchanged between channels, to solve the permutation problem.

摘要翻译： 当使用独立分量分析将混合在音频信号中的多个信号中的每个信号分离时，不用利用关于原始信号的知识或关于麦克风等的位置的信息，可以高精度地解决置换问题。 短时傅里叶变换部分从时域观测信号产生观测信号的光谱图。 信号分离部将观测信号的光谱图分离成各信号的光谱图，生成分离信号的光谱图。 置换问题解决部分从分离信号的实质上的整个谱图计算对应于置换度的比例，例如，通过使用多维概率密度函数或多维高度计算的Kullback-Leiblar信息量。 基于比例，分离信号的频谱图的每个频率段的信号在信道之间交换，以解决置换问题。

公开(公告)号：US08566094B2

公开(公告)日：2013-10-22

申请号：US13206631

申请日：2011-08-10

申请人： Katsuki Minamino ,  Atsuo Hiroe ,  Yoshinori Maeda ,  Satoshi Asakawa

发明人： Katsuki Minamino ,  Atsuo Hiroe ,  Yoshinori Maeda ,  Satoshi Asakawa

IPC分类号： G10L15/00

CPC分类号： G10L15/32 ,  G10L15/187 ,  G10L15/19

摘要： An apparatus, method and program for performing a speech recognition process utilizing contextual information that comprises an estimation of the intention of an utterance of a user. The recognition process includes calculating a pre-score based on observed contextual information according intention models which correspond to a plurality of types of intention information and combining the pre-scoring results with acoustic and linguistic scores to obtain an improved recognition or comprehension of the intent of a user utterance.

摘要翻译： 一种用于使用包括对用户的话语的意图的估计的上下文信息执行语音识别处理的装置，方法和程序。 识别过程包括基于观察到的情境信息来计算预分数，该意图模型对应于多种类型的意图信息，并将预评分结果与声学和语言得分相结合，以获得对目标的意图的改进的识别或理解 用户说话。

公开(公告)号：US20110261977A1

公开(公告)日：2011-10-27

申请号：US13071047

申请日：2011-03-24

申请人： Atsuo Hiroe

发明人： Atsuo Hiroe

IPC分类号： H04B1/00

CPC分类号： H04R3/005 ,  G10L21/0272 ,  G10L2021/02166 ,  H04R2430/03 ,  H04S7/30 ,  H04S2400/11 ,  H04S2400/15 ,  H04S2420/07

摘要： A signal processing device includes a signal transform unit which generates observation signals in the time frequency domain, and an audio source separation unit which generates an audio source separation result, and the audio source separation unit includes a first-stage separation section which calculates separation matrices for separating mixtures included in the first frequency bin data set by a learning process in which Independent Component Analysis is applied to the first frequency bin data set, and acquires a first separation result for the first frequency bin data set, a second-stage separation section which acquires a second separation result for a second frequency bin data set by using a score function in which an envelope is used as a fixed one, and executing a learning process for calculating separation matrices for separating mixtures, and a synthesis section which generates the final separation results by integrating the first and the second separation results.

摘要翻译： 一种信号处理装置，包括：时间频率域生成观测信号的信号变换部，以及产生声源分离结果的音频源分离部，音源分离部包括计算分离矩阵的第一级分离部 用于通过将独立分量分析应用于第一频率仓数据集的学习处理来分离包括在第一频率仓数据集中的混合，并且获取第一频率仓数据集的第一分离结果;第二分段部分 其通过使用其中使用包络作为固定的分数函数来获取第二频率仓数据集的第二分离结果，以及执行用于计算用于分离混合物的分离矩阵的学习处理，以及生成最终的合成部分 通过集成第一和第二分离结果的分离结果。

公开(公告)号：US20060277035A1

公开(公告)日：2006-12-07

申请号：US11421619

申请日：2006-06-01

申请人： Atsuo Hiroe ,  Keiichi Yamada

发明人： Atsuo Hiroe ,  Keiichi Yamada

IPC分类号： G10L21/00

CPC分类号： G10L21/0272

摘要： Problems of permutation can be solved with high accuracy without utilizing knowledge about original signals or information concerning positions of microphones and the like when each one of plural signals mixed in an audio signal is separated using independent component analysis. A short-time Fourier transformation section generates spectrograms of observation signals from observation signals in time domain. A signal separation section separates the spectrograms of the observation signals into spectrograms of respective signals, to generate spectrograms of separate signals. A permutation problem solution section calculates a scale corresponding to the degree of permutation, e.g., a Kullback-Leiblar information amount calculated by use of a multidimensional probability density function or multidimensional kurtosis, from substantial whole of the spectrograms of the separate signals. Based on the scale, signals at each of frequencies bin of the spectrograms of the separate signals are exchanged between channels, to solve the permutation problem.

摘要翻译： 当使用独立分量分析将混合在音频信号中的多个信号中的每个信号分离时，不用利用关于原始信号的知识或关于麦克风等的位置的信息，可以高精度地解决置换问题。 短时傅里叶变换部分从时域观测信号产生观测信号的光谱图。 信号分离部将观测信号的光谱图分离成各信号的光谱图，生成分离信号的光谱图。 置换问题解决部分从分离信号的实质上的整个谱图计算对应于置换度的比例，例如，通过使用多维概率密度函数或多维高度计算的Kullback-Leiblar信息量。 基于比例，分离信号的频谱图的每个频率段的信号在信道之间交换，以解决置换问题。

公开(公告)号：US07088853B2

公开(公告)日：2006-08-08

申请号：US10336201

申请日：2002-12-31

申请人： Atsuo Hiroe ,  Katsuki Minamino ,  Kenta Kawamoto ,  Kohtaro Sabe ,  Takeshi Ohashi

发明人： Atsuo Hiroe ,  Katsuki Minamino ,  Kenta Kawamoto ,  Kohtaro Sabe ,  Takeshi Ohashi

IPC分类号： G06K9/00 ,  G10L15/00 ,  G10L13/00

CPC分类号： G10L13/047 ,  G10L13/00 ,  G10L15/06 ,  G10L15/24 ,  G10L2015/0631

摘要： A plural number of letters or characters, inferred from the results of letter/character recognition of an image photographed by a CCD camera (20), a plural number of kana readings inferred from the letters or characters and the way of pronunciation corresponding to the kana readings are generated in an pronunciation information generating unit (150) and the plural readings obtained are matched to the pronunciation from the user acquired by a microphone (23) to specify one kana reading and the way of pronunciation (reading) from among the plural generated candidates.

摘要翻译： 从由CCD照相机（20）拍摄的图像的字母/字符识别的结果推断的多个字母或字符，从字母或字符推断的多个假名读数和对应于假名的发音方式 在发音信息生成单元（150）中产生读数，并且所获得的多个读数与由麦克风（23）获取的用户的发音相匹配，以从多个生成的多个生成单元中指定一个假名读数和发音（读取）的方式 候选人。

公开(公告)号：US09361907B2

公开(公告)日：2016-06-07

申请号：US13348260

申请日：2012-01-11

申请人： Atsuo Hiroe

发明人： Atsuo Hiroe

IPC分类号： H04R29/00 ,  G10L25/48 ,  G01S3/808 ,  G10L25/78 ,  G10L21/0216

CPC分类号： G10L25/48 ,  G01S3/8083 ,  G10L25/78 ,  G10L2021/02166

摘要： An apparatus including a direction estimation unit detecting one or more direction points indicating a sound source direction of a sound signal for each of blocks divided in a predetermined time unit, and a direction tracking unit connecting the direction points to each other between the blocks and detecting a section in which a sound is active.

摘要翻译： 一种装置，包括：方向估计单元，检测指示以预定时间单位划分的每个块的声音信号的声源方向的一个或多个方向点;以及方向跟踪单元，其将块之间的方向点彼此连接并检测 声音活动的部分。