公开(公告)号：US09922663B2

公开(公告)日：2018-03-20

申请号：US15066285

申请日：2016-03-10

Applicant: Huawei Technologies Co., Ltd.

Inventor： Rilin Chen ,  Deming Zhang

IPC: G10L21/028 ,  G10L21/0208 ,  H04R3/00 ,  G10L21/0216 ,  G10L15/22

CPC classification number: G10L21/028 ,  G10L21/0208 ,  G10L2015/228 ,  G10L2021/02087 ,  G10L2021/02166 ,  H04R3/005 ,  H04R2499/11

Abstract: A voice signal processing method and apparatus, which are used to process a voice signal collected by a microphone of a terminal in order to meet requirements of the terminal in different application modes for the voice signal generated after the processing. The method includes collecting at least two voice signals, determining a current application mode of a terminal, determining, according to the current application mode from the voice signals, voice signals corresponding to the current application mode, and performing, in a preset voice signal processing manner that matches the current application mode, beamforming processing on the corresponding voice signals.

公开(公告)号：US09792257B2

公开(公告)日：2017-10-17

申请号：US15345074

申请日：2016-11-07

Applicant: Huawei Technologies Co., Ltd.

Inventor： Deming Zhang ,  Haiting Li ,  Anisse Taleb ,  Jianfeng Xu

IPC: G06F17/14 ,  G10L19/02 ,  G10L19/022

CPC classification number: G06F17/142 ,  G06F17/141 ,  G06F17/147 ,  G10L19/0212 ,  G10L19/022

Abstract: An embodiment of the present invention discloses a data processing method, including: twiddling input data, so as to obtain twiddled data; pre-rotating the twiddled data by using a symmetric rotate factor, where the rotate factor is a·W4L2p+1, p=0, . . . , L/2−1, and a is a constant; performing a Fast Fourier (Fast Fourier Transform, FFT) transform of L/2 point on the pre-rotated data, where L is the length of the input data; post-rotating the data that has undergone the FFT transform by using a symmetric rotate factor, where the rotate factor is b·W4L2q+1, q=0, . . . , L/2−1, and b is a constant; and obtaining output data.

公开(公告)号：US20160189728A1

公开(公告)日：2016-06-30

申请号：US15066285

申请日：2016-03-10

Applicant: Huawei Technologies Co., Ltd.

Inventor： Rilin Chen ,  Deming Zhang

IPC: G10L21/028

CPC classification number: G10L21/028 ,  G10L21/0208 ,  G10L2015/228 ,  G10L2021/02087 ,  G10L2021/02166 ,  H04R3/005 ,  H04R2499/11

Abstract: A voice signal processing method and apparatus, which are used to process a voice signal collected by a microphone of a terminal in order to meet requirements of the terminal in different application modes for the voice signal generated after the processing. The method includes collecting at least two voice signals, determining a current application mode of a terminal, determining, according to the current application mode from the voice signals, voice signals corresponding to the current application mode, and performing, in a preset voice signal processing manner that matches the current application mode, beamforming processing on the corresponding voice signals.

Abstract translation: 一种语音信号处理方法和装置，其用于处理由终端的麦克风收集的语音信号，以便满足在处理之后生成的语音信号在不同应用模式下终端的要求。 该方法包括收集至少两个语音信号，确定终端的当前应用模式，根据当前应用模式根据语音信号确定与当前应用模式相对应的语音信号，并且在预设语音信号处理 与当前应用模式匹配的方式，对相应语音信号进行波束成形处理。

公开(公告)号：US20150134332A1

公开(公告)日：2015-05-14

申请号：US14598602

申请日：2015-01-16

Applicant: Huawei Technologies Co., Ltd.

Inventor： Change Liu ,  Deming Zhang

IPC: G10L15/06

CPC classification number: G10L15/06 ,  G10L15/18 ,  G10L15/187 ,  G10L2015/223

Abstract: A speech recognition method and device are disclosed. The method includes: acquiring a text file specified by a user, and extracting a command word from the text file, to obtain a command word list; comparing the command word list with a command word library, to confirm whether the command word list includes a new command word; if the command word list includes the new command word, generating a corresponding new pronunciation dictionary; merging the new language model into a language model library; and receiving speech, and performing speech recognition on the speech according to an acoustic model, a phonation dictionary, and the language model library. Command words acquired online are closely related to online content; therefore, the number of the command words is limited and far less than the number of frequently used words.

Abstract translation: 公开了一种语音识别方法和装置。 该方法包括：获取由用户指定的文本文件，并从文本文件中提取命令字，以获得命令字列表; 将命令字列表与命令字库进行比较，以确认命令字列表是否包含新的命令字; 如果命令字列表包括新命令字，则生成相应的新发音字典; 将新语言模型合并成语言模型库; 并且根据声学模型，发音字典和语言模型库对语音进行语音识别。 在线获取的指令字与在线内容密切相关; 因此，命令字的数量是有限的，远远低于经常使用的字数。

公开(公告)号：US20160173978A1

公开(公告)日：2016-06-16

申请号：US15049515

申请日：2016-02-22

Applicant: Huawei Technologies Co., Ltd.

Inventor： Haiting Li ,  Deming Zhang

IPC: H04R1/40 ,  G10L21/02

CPC classification number: H04R1/406 ,  G10L21/02 ,  G10L21/0205 ,  G10L2021/02082 ,  G10L2021/02166 ,  H04R2201/025 ,  H04R2201/405 ,  H04R2430/21

Abstract: An audio signal processing method and apparatus and a differential beamforming method and apparatus to resolve a problem that an existing audio signal processing system cannot process audio signals in multiple application scenarios at the same time. The method includes determining a super-directional differential beamforming weighting coefficient, acquiring an audio input signal and determining a current application scenario and an audio output signal, acquiring, a weighting coefficient corresponding to the current application scenario, performing super-directional differential beamforming processing on the audio input signal using the acquired weighting coefficient in order to obtain a super-directional differential beamforming signal in the current application scenario, and performing processing on the formed signal to obtain a final audio signal required by the current application scenario. By using this method, a requirement that different application scenarios require different audio signal processing manners can be met.

Abstract translation: 一种用于解决现有音频信号处理系统在多个应用场景中不能同时处理音频信号的问题的音频信号处理方法和装置以及差分波束形成方法和装置。 该方法包括：确定超方向差分波束成形加权系数，获取音频输入信号并确定当前应用场景和音频输出信号，获取与当前应用场景相对应的加权系数，执行超向导差分波束成形处理 使用所获取的加权系数的音频输入信号，以便在当前应用场景中获得超方向的差分波束形成信号，并且对形成的信号执行处理以获得当前应用场景所需的最终音频信号。 通过使用该方法，可以满足不同应用场景需要不同音频信号处理方式的要求。

公开(公告)号：US09247334B2

公开(公告)日：2016-01-26

申请号：US14554298

申请日：2014-11-26

Applicant: Huawei Technologies Co., Ltd.

Inventor： Deming Zhang ,  David Virette ,  Yue Lang

IPC: H04R21/02 ,  H04R1/08 ,  H04M1/03 ,  H04R5/027 ,  H04M1/02

CPC classification number: H04R1/08 ,  H04M1/0202 ,  H04M1/035 ,  H04R5/027 ,  H04R2499/11

Abstract: The invention relates to a portable electronic device, comprising: a housing comprising at least one hole; and at least two directional microphones mounted in the housing and placed coincidentally for stereo sound pickup, each one of the microphones defining a main sound axis and each one of the two directional microphones defining a direct sound direction and a opposite sound direction which describe opposite directions of the main sound axis, wherein the at least one hole is a common hole shared between the at least two directional microphones such that the main sound axis of the at least two directional microphones are pointing through the common hole in different directions.

Abstract translation: 本发明涉及一种便携式电子设备，包括：壳体，包括至少一个孔; 以及安装在壳体中的至少两个定向麦克风并且巧妙地放置用于立体声拾音，每个麦克风限定主声轴，并且两个定向麦克风中的每一个定义了直接声音方向和相反方向，其描述了相反的方向 的主声轴，其中所述至少一个孔是在所述至少两个定向麦克风之间共享的公共孔，使得所述至少两个定向麦克风的主声轴在不同方向上指向所述公共孔。

公开(公告)号：US09996503B2

公开(公告)日：2018-06-12

申请号：US15696091

申请日：2017-09-05

Applicant: Huawei Technologies Co., Ltd.

Inventor： Deming Zhang ,  Haiting Li ,  Anisse Taleb ,  Jianfeng Xu

IPC: G06F17/14 ,  G10L19/022 ,  G10L19/02

CPC classification number: G06F17/142 ,  G06F17/141 ,  G06F17/147 ,  G10L19/0212 ,  G10L19/022

Abstract: A method for processing an audio signal, including: sound is converted to an analog audio input signal and converted into a digital audio signal; a windowed time domain signal is obtained and then a twiddled signal is obtained; the twiddled signal is pre-rotated and then an FFT is performed; an in-place fixed rotate compensation is performed on the FFT signal and then an post-rotated is performed; a quantized signal is obtained and then wrote into a bitstream for transmitting or storing.

公开(公告)号：US20160269844A1

公开(公告)日：2016-09-15

申请号：US15163369

申请日：2016-05-24

Applicant: Huawei Technologies Co., Ltd.

Inventor： Xianghui Ding ,  Deming Zhang ,  Zhiming Long

IPC: H04S1/00

CPC classification number: H04S1/005 ,  H04R1/1041 ,  H04R5/04 ,  H04R2201/107

Abstract: A audio signal processing method includes: sampling and holding, in a time sharing manner, an audio-left channel analog signal output by an audio-left channel microphone of a stereo headset and an audio-right channel analog signal output by an audio-right channel microphone; performing analog-to-digital conversion on the audio-left channel analog signal and audio-right channel analog signal that are sampled and held, so as to obtain a corresponding audio-left channel digital signal and an audio-right channel digital signal; performing data encapsulation on the audio-left channel digital signal and the audio-right channel digital signal to obtain a serial digital signal; and receiving and demodulating, by a terminal, the serial digital signal to obtain an audio-left channel digital signal and an audio-right channel digital signal.

Abstract translation: 音频信号处理方法包括：以时分方式采样和保持由立体声头戴式耳机的音频左声道麦克风输出的音频左声道模拟信号和由声频输出的音频右声道模拟信号 通道麦克风; 对被采样和保持的音频左声道模拟信号和音频右声道模拟信号执行模数转换，以获得相应的音频左声道数字信号和音频右声道数字信号; 在音频左声道数字信号和音频右声道数字信号上执行数据封装以获得串行数字信号; 并通过终端接收和解调串行数字信号以获得音频左声道数字信号和音频右声道数字信号。

公开(公告)号：US09439019B2

公开(公告)日：2016-09-06

申请号：US14822476

申请日：2015-08-10

Applicant: Huawei Technologies Co., Ltd.

Inventor： Jingdong Chen ,  Haiting Li ,  Deming Zhang

IPC: H04R5/00 ,  H04S3/00 ,  H04R3/00 ,  H04S3/02 ,  H04R5/027 ,  H04S7/00 ,  H04W88/02 ,  H04R1/40

CPC classification number: H04S3/002 ,  H04R1/406 ,  H04R3/005 ,  H04R5/027 ,  H04R2201/401 ,  H04R2201/405 ,  H04R2203/12 ,  H04S3/02 ,  H04S7/302 ,  H04S2400/11 ,  H04S2400/15 ,  H04S2420/13 ,  H04W88/02

Abstract: A sound signal processing method and apparatus are provided that relate to the audio signal processing field. The method in the present invention includes acquiring, by a mobile terminal, sound signals from a three-dimensional sound field, where at least three microphones are disposed on the mobile terminal and one microphone is configured to receive a sound signal in at least one direction; acquiring, according to the acquired sound signals, a direction of a sound source relative to the mobile terminal; and obtaining spatial audio signals according to the direction of the sound source relative to the mobile terminal and the acquired sound signals, where the spatial audio signals are used for simulating the three-dimensional sound field. The present invention is applicable to a process of collecting and processing signals in a three-dimensional sound field surrounding a terminal.

Abstract translation: 提供了与音频信号处理领域相关的声音信号处理方法和装置。 本发明的方法包括：由移动终端获取来自三维声场的声音信号，其中在移动终端上设置有至少三个麦克风，并且一个麦克风被配置为在至少一个方向上接收声音信号 ; 根据获取的声音信号，获取声源相对于移动终端的方向; 并且根据相对于移动终端的声源的方向和获取的声音信号获得空间音频信号，其中空间音频信号用于模拟三维声场。 本发明适用于在终端周围的三维声场中采集和处理信号的处理。

公开(公告)号：US20160205263A1

公开(公告)日：2016-07-14

申请号：US15078587

申请日：2016-03-23

Applicant: Huawei Technologies Co., Ltd.

Inventor： Yuanyuan Liu ,  Deming Zhang

IPC: H04M9/08

CPC classification number: H04M9/082 ,  H04R3/02 ,  H04R2499/11

Abstract: An echo cancellation method and apparatus, where the method includes collecting, by a collection microphone, a sound signal, collecting, by a conversation microphone, a near-end speech signal, canceling an echo component in the near-end speech signal according to the sound signal, to generate an echo-canceled speech signal, and outputting the echo-canceled speech signal such that the echo cancellation effect may be improved and conversation quality may be enhanced.

Abstract translation: 一种回声消除方法和装置，其特征在于，该方法包括：收集麦克风收集声音信号，由对话麦克风收集近端语音信号，根据所述近端语音信号消除近端语音信号中的回波分量 声音信号，以产生回声消除的语音信号，并且输出回声消除的语音信号，使得可以提高回声消除效果，并且可以提高会话质量。