公开(公告)号：US20170032804A1

公开(公告)日：2017-02-02

申请号：US15113743

申请日：2014-01-24

申请人： Foundation of Soongsil University-Industry Cooperation

发明人： Myung Jin BAE ,  Sang Gil LEE ,  Geum Ran BAEK

IPC分类号： G10L25/66 ,  G10L25/93 ,  A61B5/00 ,  G10L19/12 ,  A61B5/18 ,  G10L15/16 ,  G10L25/15

CPC分类号： G10L25/66 ,  A61B5/18 ,  A61B5/4803 ,  A61B5/4845 ,  A61B2010/0009 ,  A61B2503/22 ,  G10L15/16 ,  G10L19/12 ,  G10L25/15 ,  G10L25/30 ,  G10L25/93

摘要： Disclosed is a method for determining alcohol consumption capable of analyzing alcohol consumption in a time domain by analyzing a formant slope of a voice signal, and a recording medium and a terminal for carrying out same. An terminal for determining whether a person is drunk comprises: a voice input unit for generating a voice frame by receiving a voice signal; a voiced/unvoiced sound analysis unit for determining whether a received voiced frame corresponds to a voiced sound; a formant frequency extraction unit for extracting a plurality of formant frequencies of the voice frame corresponding to the voiced sound; and an alcohol consumption determining unit for calculating a formant slope between the plurality of formant frequencies, and determining the state of alcohol consumption depending on the formant slope, thereby determining whether a person is drunk by analyzing the formant slope of an inputted voice.

摘要翻译： 公开了一种通过分析语音信号的共振峰斜率来确定能够分析时域中的酒精消耗的酒精消耗的方法，以及用于进行声音信号的记录介质和终端的方法。 用于确定人是否醉酒的终端包括：语音输入单元，用于通过接收语音信号来产生语音帧; 有声/无声音分析单元，用于确定接收到的有声帧是否对应于浊音; 共振峰频率提取单元，用于提取对应于有声声音的语音帧的多个共振峰频率; 以及酒精消耗确定单元，用于计算多个共振峰频率之间的共振峰斜率，并且根据共振峰斜率确定酒精消耗的状态，从而通过分析输入的声音的共振峰斜率来确定人是否醉酒。

公开(公告)号：US20160379627A1

公开(公告)日：2016-12-29

申请号：US15262963

申请日：2016-09-12

申请人： SPEECH MORPHING SYSTEMS, INC.

发明人： Fathy YASSA ,  Ben REAVES

IPC分类号： G10L15/183 ,  G10L15/04 ,  G10L25/93

CPC分类号： G10L15/05 ,  G10L25/09 ,  G10L25/21 ,  G10L25/78 ,  G10L25/81 ,  G10L25/84 ,  G10L25/87

摘要： Method and apparatus for segmenting speech by detecting the pauses between the words and/or phrases, and to determine whether a particular time interval contains speech or non-speech, such as a pause.

公开(公告)号：US09520141B2

公开(公告)日：2016-12-13

申请号：US13781262

申请日：2013-02-28

申请人： GOOGLE INC.

发明人： Jens Enzo Nyby Christensen ,  Simon J. Godsill ,  Jan Skoglund

IPC分类号： G10L21/02 ,  G10L25/48 ,  G10L21/0216 ,  G10L25/84 ,  G10L25/93 ,  G10L21/0208

CPC分类号： G10L25/48 ,  G10L21/02 ,  G10L21/0208 ,  G10L21/0216 ,  G10L25/84 ,  G10L25/93 ,  G10L2025/935

摘要： Provided are methods and systems for detecting the presence of a transient noise event in an audio stream using primarily or exclusively the incoming audio data. Such an approach offers improved temporal resolution and is computationally efficient. The methods and systems presented utilize some time-frequency representation of an audio signal as the basis in a predictive model in an attempt to find outlying transient noise events and interpret the true detection state as a Hidden Markov Model (HMM) to model temporal and frequency cohesion common amongst transient noise events.

摘要翻译： 提供了用于检测音频流中瞬时噪声事件的存在的方法和系统，其主要或排他地使用输入音频数据。 这种方法提供了改进的时间分辨率，并且在计算上是有效的。 所提出的方法和系统利用音频信号的一些时间频率表示作为预测模型的基础，以试图找出偏离的瞬态噪声事件，并将真实检测状态解释为隐马尔可夫模型（HMM）来模拟时间和频率 瞬态噪声事件中共同的凝聚力。

公开(公告)号：US20160284343A1

公开(公告)日：2016-09-29

申请号：US14217198

申请日：2014-03-17

申请人： Kevin M. Short ,  Brian Hone

发明人： Kevin M. Short ,  Brian Hone

IPC分类号： G10L15/02 ,  G10L25/93 ,  G10L25/24 ,  G10L25/21 ,  G10L25/18

CPC分类号： G10L15/02 ,  G10L25/03 ,  G10L25/18 ,  G10L25/21 ,  G10L25/24 ,  G10L25/93 ,  G10L2015/025

摘要： A method of renormalizing high-resolution oscillator peaks, extracted from windowed samples of an audio signal, is disclosed. Feature vectors are generated for which variations in both fundamental frequency and time duration of speech are substantially mitigated. The feature vectors may be aligned within a common coordinate space, free of those variations in frequency and time duration that occurs between speakers, and even over speech by a single speaker, to facilitate a simple and accurate determination of matches between those AFDVs generated from a sample of the audio signal and corpus AFDVs generated for known speech at the phoneme and sub-phoneme level. The renormalized feature vectors can be combined with traditional feature vectors such as MFCCs, or they can be used exclusively to identify voiced, semi-voiced and unvoiced sounds.

摘要翻译： 公开了一种从音频信号的窗口样本中提取的高分辨率振荡器峰值的重新归一化方法。 生成基本频率和语音持续时间的变化的特征向量被大大减轻。 特征向量可以在公共坐标空间内对齐，没有在扬声器之间发生的频率和持续时间的这些变化，甚至在单个扬声器的语音之间的对准，以便简单和准确地确定从一个扬声器产生的那些AFDV之间的匹配 在音素和子音素级别为已知语音生成的音频信号和语料库AFDV的样本。 重归一化特征向量可以与诸如MFCC的传统特征向量组合，或者它们可以专门用于识别有声，半声和无声的声音。

公开(公告)号：US20160277849A1

公开(公告)日：2016-09-22

申请号：US15079036

申请日：2016-03-23

申请人： ADVANCED BIONICS AG

发明人： Adam B. Strauss ,  Leonid M. Litvak

IPC分类号： H04R25/00 ,  G10L25/93

CPC分类号： H04R25/353 ,  A61N1/36032 ,  A61N1/36036 ,  G10L21/0364 ,  G10L25/93 ,  H04R25/356 ,  H04R25/50 ,  H04R25/505 ,  H04R25/606 ,  H04R2225/43 ,  H04R2225/67 ,  H04R2430/03 ,  H04R2460/01

摘要： An exemplary sound processor 1) identifies at least one frequency bin, included in a plurality of frequency bins included in a frequency spectrum of an audio signal that is presented to a cochlear implant patient, that contains spectral energy above a modified spectral envelope, 2) identifies each frequency bin that contains spectral energy below the modified spectral envelope, 3) enhances the spectral energy contained in the at least one frequency bin identified as containing spectral energy above the modified spectral envelope, and 4) compresses the spectral energy contained in each frequency bin identified as containing spectral energy below the modified spectral envelope.

摘要翻译： 示例性声音处理器1）识别包括在被呈现给耳蜗植入物患者的音频信号的频谱中的多个频率仓中的包含在修改的频谱包络之上的频谱能量的至少一个频率仓，2） 识别包含在修改的频谱包络之下的频谱能量的每个频率仓，3）增强被识别为包含在修改的频谱包络上的频谱能量的至少一个频率仓中的频谱能量，以及4）压缩每个频率中包含的频谱能量 bin被识别为包含在修改的光谱包络之下的光谱能量。

公开(公告)号：US20160189713A1

公开(公告)日：2016-06-30

申请号：US14926697

申请日：2015-10-29

申请人： HON HAI PRECISION INDUSTRY CO., LTD.

发明人： YOUNG-WAY LIU

IPC分类号： G10L15/26 ,  G10L25/93 ,  G10L15/07 ,  G06F17/24 ,  G10L17/00

CPC分类号： G06F17/248 ,  G10L15/26 ,  G10L17/00

摘要： An electronic apparatus for automatically acquiring and revising minutes of a meeting and a method thereof includes the steps of identifying one or more speakers from audio signals which are recorded during a meeting, based on pre-sampled audio signals and a voice feature table stored in a non-transitory storage medium. The audio signals are converted to text and divided into paragraphs, one paragraph being attributable to one speaker, and each speaker has a given user name. An original minutes of the meeting, based on the text and a meeting minutes template stored in the non-transitory storage medium, is prepared and revised and issued to all relevant persons.

摘要翻译： 一种用于自动获取和修改会议记录的电子装置及其方法包括以下步骤：基于预采样的音频信号和存储在会议中的语音特征表，识别在会议期间记录的音频信号中的一个或多个扬声器 非暂时性存储介质。 音频信号被转换为文本并分为段落，一个段落可归因于一个扬声器，并且每个扬声器具有给定的用户名。 根据存储在非暂时性存储介质中的文本和会议记录模板，会议的原始会议记录被准备和修改并发布给所有相关人员。

公开(公告)号：US20160189107A1

公开(公告)日：2016-06-30

申请号：US14926814

申请日：2015-10-29

申请人： HON HAI PRECISION INDUSTRY CO., LTD

发明人： YOUNG-WAY LIU

IPC分类号： G06Q10/10 ,  G10L25/93 ,  G10L15/26 ,  G10L25/87 ,  G10L17/00

CPC分类号： G06Q10/1091 ,  G10L15/26 ,  G10L25/87

摘要： A computing device for automatically acquiring and revising minutes of a meeting and a method thereof includes the steps of: identifying one or more silences or notional silences (unvoiced segments) in voice data; determining a segment as being a satisfactory unvoiced segment if the gap of silence lasts for a time period equal to or larger than a predetermined period; dividing the audio data or text representing the audio data into one or more passages of text according to the satisfactory unvoiced segment, and creating an original minutes of the meeting according to the audio data or the representative text being divided into passages and a meeting minutes template stored in the non-transitory storage medium.

摘要翻译： 一种用于自动获取和修改会议记录的计算设备及其方法包括以下步骤：识别语音数据中的一个或多个静音或无声段（无声段）; 如果所述静音间隔持续等于或大于预定周期的时间段，则确定段是令人满意的无声段; 将表示音频数据的音频数据或文本根据令人满意的无声部分分成一段或多段文本，并根据音频数据或代表性文本创建会议的原始分钟，分为段落和会议记录模板 存储在非瞬时存储介质中。

公开(公告)号：US09373342B2

公开(公告)日：2016-06-21

申请号：US14312074

申请日：2014-06-23

申请人： Nuance Communications, Inc.

发明人： Sridhar Pilli ,  Mahesh Godavarti ,  Qian-Yu Tang ,  Jose Lainez ,  Jagadeesh Balam

IPC分类号： G10L21/0364 ,  G10L19/26 ,  G10L25/12 ,  G10L19/12 ,  G10L25/78 ,  G10L25/93 ,  G10L19/16 ,  G10L25/21

CPC分类号： G10L21/0364 ,  G10L19/12 ,  G10L19/173 ,  G10L19/26 ,  G10L25/12 ,  G10L25/21 ,  G10L25/78 ,  G10L25/93

摘要： The present disclosure is directed towards a method for speech intelligibility. The method may include receiving, at one or more computing devices, a first speech input from a first user and performing voice activity detection upon the first speech input. The method may also include analyzing a spectral tilt associated with the first speech input, wherein analyzing includes computing an impulse response of a linear predictive coding (“LPC”) synthesis filter in a linear pulse code modulation (“PCM”) domain and wherein the one or more computing devices includes an adaptive high pass filter configured to recalculate one or more linear prediction coefficients.

摘要翻译： 本公开涉及一种用于语音清晰度的方法。 该方法可以包括在一个或多个计算设备处接收来自第一用户的第一语音输入并且在第一语音输入时执行语音活动检测。 该方法还可以包括分析与第一语音输入相关联的频谱倾斜，其中分析包括在线性脉码调制（“PCM”）域中计算线性预测编码（“LPC”）合成滤波器的脉冲响应，并且其中 一个或多个计算设备包括被配置为重新计算一个或多个线性预测系数的自适应高通滤波器。

公开(公告)号：US20160155441A1

公开(公告)日：2016-06-02

申请号：US14670149

申请日：2015-03-26

申请人： TATA CONSULTANCY SERVICES LTD.

发明人： Ashish Panda ,  Sunil Kumar Kopparapu

IPC分类号： G10L15/20 ,  G10L25/93

CPC分类号： G10L25/51 ,  G10L15/06

摘要： The present disclosure envisages a computer implemented system for identifying significant speech frames within speech signals for facilitating speech recognition. The system receives an input speech signal having a plurality of feature vectors which is passed through a spectrum analyzer. The spectrum analyzer divides the input speech signal into a plurality of speech frames and computes a spectral magnitude of each of the speech frames. There is provided a suitability engine which is enabled to compute a suitability measure for each of the speech frames corresponding to spectral flatness measure (SFM), energy normalized variance (ENV), entropy, signal-to-noise ratio (SNR) and similarity measure. The suitability engine further computes a weighted suitability measure for each of the speech frames.

摘要翻译： 本公开设想一种计算机实现的系统，用于识别语音信号内的有效语音帧以便于语音识别。 系统接收具有通过频谱分析仪的多个特征向量的输入语音信号。 频谱分析仪将输入语音信号划分为多个语音帧，并计算每个语音帧的频谱幅度。 提供了一种适合性引擎，其能够针对与频谱平坦度测量（SFM），能量归一化方差（ENV），熵，信噪比（SNR）和相似性度量相对应的每个语音帧计算适合性度量 。 适合性引擎进一步计算每个语音帧的加权适合性度量。

公开(公告)号：US09305567B2

公开(公告)日：2016-04-05

申请号：US13827894

申请日：2013-03-14

申请人： QUALCOMM Incorporated

发明人： Erik Visser ,  Lae-Hoon Kim ,  Jongwon Shin ,  Yinyi Guo ,  Sang-Uk Ryu ,  Andre Gustavo P. Schevciw

IPC分类号： G10L21/0208 ,  G10L15/20 ,  G10L21/0316 ,  G10L25/93 ,  G10L21/0216

CPC分类号： G10L21/0208 ,  G10L15/20 ,  G10L21/0316 ,  G10L25/93 ,  G10L2021/02165

摘要： A method for signal level matching by an electronic device is described. The method includes capturing a plurality of audio signals from a plurality of microphones. The method also includes determining a difference signal based on an inter-microphone subtraction. The difference signal includes multiple harmonics. The method also includes determining whether a harmonicity of the difference signal exceeds a harmonicity threshold. The method also includes preserving the harmonics to determine an envelope. The method further applies the envelope to a noise-suppressed signal.

摘要翻译： 描述了一种由电子设备进行信号电平匹配的方法。 该方法包括从多个麦克风中捕获多个音频信号。 该方法还包括基于麦克风间减法确定差分信号。 差分信号包括多个谐波。 该方法还包括确定差分信号的谐波是否超过谐波阈值。 该方法还包括保存谐波以确定信封。 该方法进一步将包络应用于噪声抑制信号。