公开(公告)号：US09711165B2

公开(公告)日：2017-07-18

申请号：US14984089

申请日：2015-12-30

申请人： AUDIONAMIX

发明人： Romain Hennequin

IPC分类号： G10L21/00 ,  H04B3/20 ,  G10L21/0308 ,  G10L19/02 ,  G10L21/02 ,  G10L21/0208

CPC分类号： G10L21/0308 ,  G10L19/02 ,  G10L21/0205 ,  G10L21/0208 ,  G10L2021/02082

摘要： Processes are described herein for transforming an audio mixture for which a specific component is affected by reverberation, into a specific dry component (i.e. unaffected by the reverberation) and a background component. In the process described herein, the long-term effects of reverberation are explicitly taken into account by modelling the spectrogram of the specific component as the result of a matrix convolution along time between the spectrogram of the specific dry component and a reverberation matrix. Parameters of the model are estimated iteratively by minimizing a cost-function measuring the divergence between the spectrogram of the mixture signal and the model of the spectrogram of the mixture signal.

公开(公告)号：US09679579B1

公开(公告)日：2017-06-13

申请号：US15186993

申请日：2016-06-20

申请人： Google Inc.

发明人： Richard Francis Lyon ,  Ron Weiss ,  Thomas Chadwick Walters

IPC分类号： G10L21/028 ,  G10L21/0356 ,  G10L25/51 ,  G10L21/0308 ,  G10L21/0388

CPC分类号： G10L21/0356 ,  G10L21/02 ,  G10L21/0272 ,  G10L21/028 ,  G10L21/0308 ,  G10L21/0388 ,  G10L25/51

摘要： Systems and methods facilitating removal of content from audio files are described. A method includes identifying a sound recording in a first audio file, identifying a reference file having at least a defined level of similarity to the sound recording, and processing the first audio file to remove the sound recording and generate a second audio file. In some embodiments, winner-take-all coding and Hough transforms are employed for determining alignment and rate adjustment of the reference file in the first audio file. After alignment, the reference file is filtered in the frequency domain to increase similarity between the reference file and the sound recording. The frequency domain representation (FR) of the filtered version is subtracted from the FR first audio and the result converted to a time representation of the second audio file. In some embodiments, spectral subtraction is also performed to generate a further improved second audio file.

公开(公告)号：US20170025133A1

公开(公告)日：2017-01-26

申请号：US15191932

申请日：2016-06-24

申请人： NANNING FUGUI PRECISION INDUSTRIAL CO., LTD. ,  HON HAI PRECISION INDUSTRY CO., LTD.

发明人： LIAN LIU

IPC分类号： G10L21/0224 ,  G10L21/0364 ,  G10L21/0308

CPC分类号： G10L21/0224 ,  G10L2021/02165

摘要： A noise elimination circuit of particular application in enhancing vocal clarity in a teleconference includes a first voice processing circuit, a second voice processing circuit, and a subtracter. The first voice processing circuit receives and processes a first voice from a first microphone and the second voice processing circuit receives and processes the same voice from a second microphone (second voice). The first voice and the second voice include voice signals and noises. The subtracter is electrically connected to the two voice processing circuits to receive the first voice and the second voice respectively processed by the first voice processing circuit and the second voice processing circuit. The subtracter substracts the second voice from the first voice, and outputs a clear voice from which noise has been eliminated.

摘要翻译： 特别应用于提高电话会议中的声音清晰度的噪声消除电路包括第一语音处理电路，第二语音处理电路和减法器。 第一语音处理电路接收并处理来自第一麦克风的第一声音，并且第二语音处理电路从第二麦克风接收并处理相同的声音（第二声音）。 第一个声音和第二个声音包括语音信号和噪音。 减法器电连接到两个语音处理电路，以接收由第一语音处理电路和第二语音处理电路分别处理的第一语音和第二语音。 减法器从第一个声音中减去第二个声音，并输出一个清除的声音，从中消除了噪声。

公开(公告)号：US20160358619A1

公开(公告)日：2016-12-08

申请号：US14732715

申请日：2015-06-06

申请人： Apple Inc.

发明人： Sean A. Ramprashad ,  Harvey D. Thornburg ,  Arvindh Krishnaswamy ,  Aram M. Lindahl

IPC分类号： G10L21/0308 ,  G10L15/34 ,  G10L15/16

CPC分类号： G10L15/34 ,  G10L15/16 ,  G10L15/20 ,  G10L2015/022 ,  G10L2021/02166

摘要： A speech recognition system for resolving impaired utterances can have a speech recognition engine configured to receive a plurality of representations of an utterance and concurrently to determine a plurality of highest-likelihood transcription candidates corresponding to each respective representation of the utterance. The recognition system can also have a selector configured to determine a most-likely accurate transcription from among the transcription candidates. As but one example, the plurality of representations of the utterance can be acquired by a microphone array, and beamforming techniques can generate independent streams of the utterance across various look directions using output from the microphone array.

摘要翻译： 用于解决受损话语的语音识别系统可以具有语音识别引擎，其被配置为接收话语的多个表示，并且同时确定与发音的每个相应表示相对应的多个最高似然转录候选。 识别系统还可以具有选择器，其被配置为确定转录候选物中最可能的精确转录。 作为一个示例，可以通过麦克风阵列来获取话音的多个表示，并且波束成形技术可以使用来自麦克风阵列的输出在各种外观方向上产生独立的话音流。

公开(公告)号：US20160275964A1

公开(公告)日：2016-09-22

申请号：US15074579

申请日：2016-03-18

申请人： Electronics and Telecommunications Research Institute

发明人： Hyun Woo KIM ,  Ho Young JUNG ,  Jeon Gue PARK ,  Yun Keun LEE

IPC分类号： G10L21/0232 ,  G10L15/14 ,  G10L19/26 ,  G10L21/0308 ,  G10L25/24

CPC分类号： G10L15/20 ,  G10L15/02

摘要： A feature compensation apparatus includes a feature extractor configured to extract corrupt speech features from a corrupt speech signal with additive noise that consists of two or more frames; a noise estimator configured to estimate noise features based on the extracted corrupt speech features and compensated speech features; a probability calculator configured to calculate a correlation between adjacent frames of the corrupt speech signal; and a speech feature compensator configured to generate compensated speech features by eliminating noise features of the extracted corrupt speech features while taking into consideration the correlation between adjacent frames of the corrupt speech signal and the estimated noise features, and to transmit the generated compensated speech features to the noise estimator.

摘要翻译： 特征补偿装置包括特征提取器，其被配置为从具有由两个或更多个帧组成的附加噪声的损坏语音信号中提取损坏的语音特征; 噪声估计器，被配置为基于所提取的损坏的语音特征和补偿的语音特征来估计噪声特征; 概率计算器，被配置为计算所述损坏语音信号的相邻帧之间的相关性; 以及语音特征补偿器，被配置为通过消除所提取的损坏的语音特征的噪声特征来产生补偿的语音特征，同时考虑到损坏的语音信号的相邻帧与估计的噪声特征之间的相关性，并且将生成的补偿语音特征发送到 噪声估计器。

公开(公告)号：US09373320B1

公开(公告)日：2016-06-21

申请号：US13972262

申请日：2013-08-21

申请人： Google Inc.

发明人： Richard Francis Lyon ,  Ron Weiss ,  Thomas Chadwick Walters

IPC分类号： G10L21/028 ,  G10L13/033 ,  G10L21/02 ,  G10L21/0272 ,  G10L21/0308

CPC分类号： G10L21/0356 ,  G10L21/02 ,  G10L21/0272 ,  G10L21/028 ,  G10L21/0308 ,  G10L21/0388 ,  G10L25/51

摘要： Systems and methods facilitating removal of content from audio files are described. A method includes identifying a sound recording in a first audio file, identifying a reference file having at least a defined level of similarity to the sound recording, and processing the first audio file to remove the sound recording and generate a second audio file. In some embodiments, winner-take-all coding and Hough transforms are employed for determining alignment and rate adjustment of the reference file in the first audio file. After alignment, the reference file is filtered in the frequency domain to increase similarity between the reference file and the sound recording. The frequency domain representation (FR) of the filtered version is subtracted from the FR first audio and the result converted to a time representation of the second audio file. In some embodiments, spectral subtraction is also performed to generate a further improved second audio file.

摘要翻译： 描述了有助于从音频文件中移除内容的系统和方法。 一种方法包括：识别在第一音频文件中的声音记录，识别具有至少与所述声音记录的确定的相似度水平的参考文件，以及处理所述第一音频文件以去除所述声音记录并生成第二音频文件。 在一些实施例中，采用获胜者全部编码和霍夫变换来确定第一音频文件中的参考文件的对准和速率调整。 对齐后，参考文件在频域中进行过滤，以增加参考文件和录音之间的相似性。 从FR第一音频中减去滤波版本的频域表示（FR），并将结果转换为第二音频文件的时间表示。 在一些实施例中，还执行频谱减法以产生进一步改进的第二音频文件。

公开(公告)号：US20160155457A1

公开(公告)日：2016-06-02

申请号：US15019242

申请日：2016-02-09

申请人： TELEFONAKTIEBOLAGET L M ERICSSON (PUBL)

发明人： Stefan Bruhn

IPC分类号： G10L25/84 ,  G10L21/0308 ,  G10L19/06

CPC分类号： G10L19/087 ,  G10L19/012 ,  G10L19/06 ,  G10L19/26 ,  G10L21/02 ,  G10L21/0308 ,  G10L25/84 ,  G10L2021/02168

摘要： In a method for coding of information for enhancing a background noise representation, voice activity of an input speech signal is determined. A noisiness parameter is determined for an inactive speech signal, wherein the noisiness parameter is based on a ratio of prediction gains of two Linear Predictive Coder (LPC) prediction filters with different orders. The noisiness parameter is quantized, and the quantized noisiness parameter is encoded for transmission.

摘要翻译： 在用于增强背景噪声表示的信息的编码方法中，确定输入语音信号的语音活动。 对于无效语音信号确定噪声参数，其中噪声参数基于具有不同阶数的两个线性预测编码器（LPC）预测滤波器的预测增益的比率。 噪声参数被量化，量化噪声参数被编码用于传输。

公开(公告)号：US20160155434A1

公开(公告)日：2016-06-02

申请号：US14951476

申请日：2015-11-24

申请人： Gregory C. Burnett ,  Eric F. Breitfeller

发明人： Gregory C. Burnett ,  Eric F. Breitfeller

IPC分类号： G10K11/178 ,  G10L21/02 ,  H04R1/46 ,  G10L21/0308

CPC分类号： G10L21/0205 ,  G10K11/178 ,  G10K2210/30232 ,  G10K2210/3028 ,  G10K2210/3045 ,  G10L19/0204 ,  G10L21/02 ,  G10L21/0208 ,  G10L21/0308 ,  G10L25/78 ,  G10L2021/02082 ,  G10L2021/02161 ,  G10L2021/02165 ,  G10L2021/02168 ,  H04R1/46

摘要： Acoustic noise suppression is provided in multiple-microphone systems using Voice Activity Detectors (VAD). A host system receives acoustic signals via multiple microphones. The system also receives information on the vibration of human tissue associated with human voicing activity via the VAD. In response, the system generates a transfer function representative of the received acoustic signals upon determining that voicing information is absent from the received acoustic signals during at least one specified period of time. The system removes noise from the received acoustic signals using the transfer function, thereby producing a denoised acoustic data stream.

摘要翻译： 在使用语音活动检测器（VAD）的多麦克风系统中提供噪声抑制。 主机系统通过多个麦克风接收声信号。 该系统还通过VAD接收与人体发声活动相关的人体组织的振动信息。 作为响应，当在至少一个指定的时间段内确定在接收到的声信号中不存在发声信息时，系统产生代表所接收的声信号的传递函数。 该系统使用传递函数从接收的声信号中去除噪声，从而产生去噪的声学数据流。

公开(公告)号：US20160099008A1

公开(公告)日：2016-04-07

申请号：US14874641

申请日：2015-10-05

申请人： Oticon A/S

发明人： Thomas BARKER ,  Tuomas VIRTANEN ,  Niels Henrik PONTOPPIDAN

IPC分类号： G10L21/0308 ,  H04R25/00 ,  G10L21/0224 ,  G10L15/20

CPC分类号： H04R25/505 ,  G10L21/028 ,  H04R2225/43 ,  H04S1/005

摘要： The application relates to a hearing device comprising a) an input unit for delivering a time varying electric input signal representing an audio signal comprising at least two sound sources, b) a cyclic analysis buffer unit of length A adapted for storing the last A audio samples, c) a cyclic synthesis buffer unit of length, where L is smaller than A, adapted for storing the last L audio samples, which are intended to be separated in individual sound sources, d) a database having stored recorded sound examples from said at least two sound sources, each entry in the database being termed an atom, the atoms originating from audio samples from first and second buffers corresponding in size to said synthesis and analysis buffer units, where for each atom, the audio samples from the first buffer overlaps with the audio samples from the second buffer, and where atoms originating from the first buffer constitute a reconstruction dictionary, and where atoms originating from the second buffer constitute an analysis dictionary. The application further relates to a method of separating audio sources, and e) a sound source separation unit for separating said electric input signal to provide separated signals representing said at least two sound sources, the sound source separation unit being configured to determine the most optimal representation (W) of the last A samples given the atoms in the analysis dictionary of the database, and to generate said at least two sound sources by combining atoms in the reconstruction dictionary of the database using the optimal representation (W). The invention may e.g. be used for hearing devices, e.g. hearing aids, headsets, ear phones, active ear protection systems, handsfree telephone systems, mobile telephones, teleconferencing systems, public address systems, classroom amplification systems, etc.

摘要翻译： 本申请涉及一种听力设备，包括：a）输入单元，用于传送表示包括至少两个声源的音频信号的时变电输入信号; b）长度为A的循环分析缓冲单元，适于存储最后A个音频采样 c）长度小于A的循环合成缓冲单元，其适于存储要在各个声源中分离的最后L个音频样本，d）具有存储的记录声音示例的数据库， 至少两个声源，数据库中的每个条目被称为原子，源自来自第一和第二缓冲器的音频样本的原子大小对应于所述合成和分析缓冲器单元，其中对于每个原子，来自第一缓冲器的音频样本重叠 来自第二缓冲器的音频样本，并且其中源自第一缓冲器的原子构成重构词典，并且其中源自第二缓冲器的原子 构成分析词典。 该应用还涉及一种分离音频源的方法，以及e）声源分离单元，用于分离所述电输入信号以提供表示所述至少两个声源的分离信号，声源分离单元被配置为确定最优 给出数据库分析词典中的原子的最后A个样本的表示（W），并且通过使用最优表示（W）通过组合数据库的重建词典中的原子来生成所述至少两个声源。 本发明可以例如 用于听力装置，例如 助听器，耳机，耳机，主动耳朵保护系统，免提电话系统，移动电话，电话会议系统，公共广播系统，教室放大系统等。

公开(公告)号：US09113240B2

公开(公告)日：2015-08-18

申请号：US12405057

申请日：2009-03-16

申请人： Dinesh Ramakrishnan ,  Song Wang

发明人： Dinesh Ramakrishnan ,  Song Wang

IPC分类号： H04B15/00 ,  H04R3/00 ,  G10L21/028 ,  G10L21/0216 ,  G10L21/0308 ,  H04R29/00

CPC分类号： H04R3/005 ,  G10L21/028 ,  G10L21/0308 ,  G10L2021/02165 ,  G10L2021/02166 ,  H04R29/006 ,  H04R2420/07 ,  H04R2430/03 ,  H04R2430/20 ,  H04R2499/11

摘要： Signal processing solutions take advantage of microphones located on different devices and improve the quality of transmitted voice signals in a communication system. With usage of various devices such as Bluetooth headsets, wired headsets and the like in conjunction with mobile handsets, multiple microphones located on different devices are exploited for improving performance and/or voice quality in a communication system. Audio signals are recorded by microphones on different devices and processed to produce various benefits, such as improved voice quality, background noise reduction, voice activity detection and the like.

摘要翻译： 信号处理解决方案利用位于不同设备上的麦克风，并提高通信系统中传输的语音信号的质量。 随着诸如蓝牙耳机，有线耳机等各种设备的使用，与移动手机结合使用，位于不同设备上的多个麦克风被用来改善通信系统中的性能和/或语音质量。 音频信号由不同设备上的麦克风记录，并被处理以产生各种益处，例如改进的语音质量，背景噪声降低，语音活动检测等。