WO2020151133A1 - 一种分布式麦克风阵列拾音***及方法 - Google Patents

一种分布式麦克风阵列拾音***及方法 Download PDF

Info

Publication number
WO2020151133A1
WO2020151133A1 PCT/CN2019/086768 CN2019086768W WO2020151133A1 WO 2020151133 A1 WO2020151133 A1 WO 2020151133A1 CN 2019086768 W CN2019086768 W CN 2019086768W WO 2020151133 A1 WO2020151133 A1 WO 2020151133A1
Authority
WO
WIPO (PCT)
Prior art keywords
microphone
microphone array
distributed
sound
signal
Prior art date
Application number
PCT/CN2019/086768
Other languages
English (en)
French (fr)
Inventor
郑能恒
丁俊豪
文衍晖
Original Assignee
深圳大学
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 深圳大学 filed Critical 深圳大学
Publication of WO2020151133A1 publication Critical patent/WO2020151133A1/zh

Links

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04RLOUDSPEAKERS, MICROPHONES, GRAMOPHONE PICK-UPS OR LIKE ACOUSTIC ELECTROMECHANICAL TRANSDUCERS; DEAF-AID SETS; PUBLIC ADDRESS SYSTEMS
    • H04R1/00Details of transducers, loudspeakers or microphones
    • H04R1/08Mouthpieces; Microphones; Attachments therefor

Definitions

  • the present invention relates to the field of speech recognition technology, and more particularly to a distributed microphone array pickup system and method.
  • the first step is to allow the device to obtain clear voice signals, that is, to make the device hear better. clear.
  • the main purpose of the present invention is to provide a distributed microphone array pickup system and method, which aims to solve the problem of poor pickup effect of voice assistants in 5 existing technologies, and the devices are bound to the voice assistant.
  • the first aspect of the present invention provides a distributed microphone array pickup system, including: a distributed microphone array for collecting sound information indoors; a sound source location processing module for locating speakers The position of the sound source emitting the sound information; a voice processing module for processing according to the sound source position and the sound information collected by the distributed microphone array to obtain a clean and smooth voice signal; a microphone data collection module for collecting the distributed microphone The array is transmitted after being processed by the voice processing module And synchronize the signal emitted by the distributed microphone array to a data acquisition terminal, and the data acquisition terminal controls the device through the signal emitted by the distributed microphone array.
  • the distributed microphone array includes microphones installed on the indoor roof and four walls; by adjusting the positions of the microphones, the microphones on the walls and the microphones on the roof are connected to each other. In space, it appears in the shape of a quadrangular pyramid.
  • the sound source localization processing module includes: a microphone coordinate determination unit, configured to set the three-dimensional coordinates of one of the reference microphones in the distributed microphone array, and set the three-dimensional coordinates according to the set three-dimensional coordinates The three-dimensional coordinate system and the three-dimensional coordinates of other microphones in the distributed microphone array; a time delay estimation unit for estimating the time delay of each microphone relative to the reference microphone according to the cross-correlation function of the signals received by the microphones; sound source position coordinates The determining unit is configured to calculate the coordinates of the sound source position according to the time, delay, sound velocity between the sound source reaching each microphone, and the coordinates of each microphone.
  • the voice processing module includes: a noise reduction unit, configured to reduce noise on the sound information collected by the distributed microphone array to obtain a clean voice signal; and a weighted average unit, configured to combine each microphone The signals are delayed and aligned, and then weighted and averaged to obtain a clean and smooth speech signal
  • the noise reduction unit includes: a main microphone signal setting subunit, which is used to set the signal of the main microphone to be composed of a human voice signal and a noise signal; a microphone determination subunit, which is used to determine the sound source The distance to each microphone determines that the microphone closest to the sound source is the main microphone, and the remaining microphones are secondary microphones, and the microphone farthest from the sound source is the noise signal reference microphone; the noise estimation subunit is used to use the noise signal reference microphone The initial silent signal is used as a noise estimate, and the noise estimate is used as the noise signal of the distributed microphone array; a power spectrum calculation subunit, configured to set the subunit and the noise estimation subunit according to the main microphone signal Calculate the power spectrum of the main microphone signal; a factor introduction subunit, used to introduce an over-subtraction factor and a spectrum lower limit compensation factor into the power spectrum calculation subunit; an inverse Fourier transform subunit, used to pass Fourier The inverse transform transforms the signal of the power spectrum calculation subunit that introduces the factor introduction
  • the weighted average unit includes: a time delay estimation subunit, configured to estimate the time delay of each microphone relative to the reference microphone according to the cross-correlation function of the signal received by Kefeng; The sub-unit is used to delay and align the voice signals received by all microphones according to the delay time of each signal, so as to perform a weighted average of each voice signal that has undergone noise reduction processing according to the distance from the sound source position to obtain a clean Smooth voice signal.
  • the microphone data collection module includes: a data collection unit, configured to collect sound information collected by all microphones in the distributed microphone array and processed by the voice processing module; a data sending unit, configured to transfer the The sound information collected by the data collection unit is sent to the background data collector; the equipment control unit is used to connect to the equipment and control the equipment according to the sound information received by the data collector.
  • the second aspect of the present invention provides a distributed microphone array pickup method, including: collecting sound information indoors through the distributed microphone array; locating the speaker's sound source position; collecting according to the sound source position and the distributed microphone array Process the sound information of the distributed microphone array to obtain a clean and smooth voice signal; collect the signals emitted by the distributed microphone array, and synchronize the signals emitted by each microphone in the distributed microphone array to a data collection terminal, and the data collection terminal passes all The signal control equipment emitted by the distributed microphone array.
  • a third aspect of the present invention provides an electronic device, including: a memory, a processor, and a computer program stored on the memory and capable of running on the processor, wherein the processor executes all When the computer program is described, the above method is realized.
  • a fourth aspect of the present invention provides a computer-readable storage medium on which a computer program is stored, which is characterized in that, when the computer program is executed by a processor, the foregoing method is implemented.
  • the present invention provides a distributed microphone array pickup system and method.
  • the beneficial effects are: Array layout of multiple microphones indoors through a distributed structure, so that there are multiple locations in the room that can collect voice information, thereby It eliminates the need for users to go to a specific device for voice interaction and emit sound information, thereby improving the sound pickup effect; and binding multiple devices with the microphone data collection module, that is, combining a distributed microphone array After the transmitted signal is synchronized to a data collection terminal, the system can control multiple devices through voice information without having to install an independent pickup module and voice assistant for each device, thereby reducing waste of resources.
  • FIG. 1 is a schematic block diagram of the structure of a distributed microphone array pickup system according to an embodiment of the present invention
  • FIG. 2 is a schematic block diagram of the structure of a sound source localization processing module of a distributed microphone array sound pickup system according to an embodiment of the present invention
  • FIG. 3 is a schematic block diagram of the structure of a voice processing module of a distributed microphone array pickup system according to an embodiment of the present invention
  • FIG. 4 is a schematic block diagram of a structure of a noise reduction unit of a distributed microphone array sound pickup system according to an embodiment of the present invention
  • FIG. 5 is a schematic block diagram of the structure of an electronic device according to an embodiment of the present invention.
  • FIG. 1 is a distributed microphone array pickup system, including: a distributed microphone array 1, used to collect sound information indoors; and a sound source localization processing module 2, used to locate the speaker emitting sound information The position of the sound source; the voice processing module 3 is used to process the sound information collected by the sound source position and the distributed microphone array 1 to obtain a clean and smooth voice signal; the microphone data collection module 4 is used to collect the distributed microphone array 1 The transmitted signal is processed by the voice processing module 3, and the signal transmitted by the distributed microphone array 1 is synchronized to a data acquisition terminal, and the data acquisition terminal controls the equipment through the signal transmitted by the distributed microphone array 1.
  • the distributed microphone array 1 includes microphones installed on the indoor roof and four walls; by adjusting the position of each microphone, the microphones on the walls and the microphones on the roof are connected to each other.
  • the line takes the shape of a quadrangular pyramid in space.
  • the sound source localization processing module 2 includes: a microphone coordinate determination unit 21, configured to set the three-dimensional coordinates of one of the reference microphones in the distributed microphone array 1, and set the three-dimensional coordinates according to the set three-dimensional coordinates The coordinate system and the three-dimensional coordinates of other microphones in the distributed microphone array 1;
  • the time delay estimation unit 22 is used to estimate the time delay of each microphone relative to the reference microphone according to the cross-correlation function of the signals received by the microphones;
  • the sound source position coordinates are determined
  • the unit 23 is configured to calculate the coordinates of the sound source position according to the time, delay, sound speed between the sound source reaching each microphone, and the coordinates of each microphone.
  • the dimensions of a rectangular room along the x-axis, y-axis, and z-axis directions in the three-dimensional space coordinate system are set as a, b, and c, respectively, and a microphone is formed on each of the four walls and ceiling of the room to form a distribution Microphone array 1, set the coordinates of one of the microphones as
  • the coordinates of the other microphones are
  • the time delay between the sound signal reaching the mth microphone of the distributed microphone array and the first microphone from the sound source position can be calculated as:
  • Formula 3 is obtained by subtracting Formula 1 from Formula 2, and Formula 3 is expressed as follows: [0042]
  • the distance from the sound source position to the m-th microphone can be expressed as:
  • the distance from the sound source position to the first microphone can be expressed as:
  • TDOA time delay estimation
  • GCC-PHAT generalized cross-correlation
  • the speech processing module 3 includes: a noise reduction unit 31, configured to reduce the noise of the sound information collected by the distributed microphone array IJ1 to obtain a clean speech signal; a weighted average unit 32, It is used to align the delay of each microphone signal, and then perform a weighted average on it to obtain a clean and smooth voice signal.
  • the noise reduction unit 31 includes: a main microphone signal setting sub-unit 311, which is used to set the signal of the main microphone to be composed of human voice signals and noise signals; a microphone determining sub-unit 312, which is used to arrive at various The distance of the microphone determines that the microphone closest to the sound source is the main microphone, and the remaining microphones are secondary microphones, and the microphone furthest from the sound source is determined to be the noise signal reference microphone; noise estimation subunit 3 13 is used to use the noise signal reference microphone The initial silent signal is used as the noise estimate, and the noise estimate is used as the noise signal of the distributed microphone array; the power spectrum calculation subunit 314 is configured to calculate the signal of each microphone signal according to the main microphone signal setting subunit 311 and the noise estimation subunit 313 Power spectrum; factor introduction subunit 315, used to introduce over-subtraction factor and spectrum lower limit compensation factor in power spectrum calculation subunit 314; inverse Fourier transform subunit 316, used to introduce factors through inverse Fourier transform The signal of the power spectrum calculation subunit 314 of
  • the main microphone signal is set
  • Noise signal let the noise signal refer to the initial silent signal of the microphone as . Since 5 microphones work at the same time, the noise signal received by all microphones is the same. Use the noise signal to refer to the initial silent section of the microphone as the noise estimation, so
  • the power spectrum may be negative due to oversubtraction, so an oversubtraction factor is introduced
  • the weighted average unit 32 includes: a time delay estimation subunit 321, configured to estimate the time delay of each microphone relative to the reference microphone according to the cross-correlation function of the signal received by the microphone; and a delay alignment subunit 322, configured to Delay alignment of the voice signals received by all microphones according to the delay time of each signal In this way, the weighted average of the voice signals that have undergone noise reduction processing is performed according to the distance from the sound source position to obtain a clean and smooth voice signal.
  • each microphone channel can obtain a relatively clean voice signal.
  • the position of the speaker does not change in a short period of time. Delay, align the five microphone signals with delay, and then perform a weighted average according to their respective distances from the speaker to obtain a clean and smooth voice signal.
  • the microphone data collection module 4 includes: a data collection unit 41 for collecting sound signals collected by all microphones in the distributed microphone array 1 and processed by the voice processing module 3; a data sending unit 42 for sending the data collection unit 41 The collected sound information is sent to the background data collector; the device control unit 43 is used to connect to the device and control the device according to the sound information received by the data collector.
  • the distributed microphone array 1 is connected to the data collection unit 41 through a wireless network, and after the voice signal received by the microphone is processed by the voice processing module 3, a clean and smooth voice signal is obtained, and then a clean and smooth voice signal is obtained through the wireless network.
  • the voice signal is transmitted to the data collection unit 41, and the data sending unit 42 transmits the voice signal received by the data collection unit 41 to the background data collector.
  • the device control unit 43 After the device control unit 43 is connected to the device, it can be controlled by the voice signal received by the data collector equipment.
  • a unified device control unit 43 is used as a central processing platform.
  • a common hardware input interface can be provided for each intelligent voice interaction device, so that the system can control multiple devices through voice information without having to install a pickup module and voice assistant for each device, which reduces the waste of resources.
  • the present application also provides a method for collecting sound from a distributed microphone array, including: collecting sound information indoors through the distributed microphone array; locating the speaker's sound source position; according to the sound source position and the sound collected by the distributed microphone array The information is processed to obtain a clean and smooth voice signal; collect the signal emitted by the distributed microphone array, and synchronize the signal emitted by the distributed microphone array microphone to a data acquisition terminal, and the data acquisition terminal controls the equipment through the signal emitted by the distributed microphone array .
  • the distributed microphone array includes microphones arranged on the indoor roof and four walls; by adjusting the position of each microphone, the microphones installed on the walls and the microphones installed on the roof are connected to each other and present a quadrangular pyramid shape in space. .
  • Locating the speaker's sound source position includes: setting the three-dimensional coordinates of one of the reference microphones in the distributed microphone array, and setting the three-dimensional coordinate system and the three-dimensional coordinates of other microphones in the distributed microphone array according to the set three-dimensional coordinates ; Estimate the time delay of each microphone relative to the reference microphone according to the cross-correlation function of the signal received by the microphone; calculate the coordinates of the sound source position according to the time, delay, sound velocity and the coordinates of each microphone between the sound source reaching each microphone.
  • Processing according to the sound source position and the sound information collected by the distributed microphone array to obtain a clean and smooth voice signal includes: denoising the sound information collected by the distributed microphone array to obtain a clean voice signal; The voice signals of each microphone in the array undergo delay alignment and weighted average to obtain a clean and smooth voice signal.
  • Denoising the sound information collected by the distributed microphone array to obtain a clean voice signal includes: setting the signal of the main microphone to be composed of a voice signal and a noise signal emitted by a person; and determining the distance according to the distance from the sound source to each microphone The microphone with the closest sound source is the main microphone, and the remaining microphones are secondary microphones.
  • the microphone farthest from the sound source is determined as the noise signal reference microphone; the initial silent signal of the noise signal reference microphone is used as the noise estimate, and the noise estimate is used as the distribution
  • the noise signal of a microphone array calculate the power spectrum of the main microphone signal according to the main microphone signal setting sub-unit and the noise estimation sub-unit; In the process of calculating the power spectrum of the main microphone signal, an over-subtraction factor and a spectrum lower limit compensation factor are introduced; The inverse inner transform transforms the signal of the power spectrum calculation subunit that introduces the factor introduction unit into a time domain signal, thereby obtaining a clean speech signal.
  • Processing the voice signal of each microphone in the distributed microphone array to obtain a clean and smooth voice signal includes: estimating the time delay of each microphone relative to the reference microphone according to the cross-correlation function of the signal received by Kefeng; The delay time of each channel of signals aligns the delays of the voice signals received by all microphones, so that the voice signals of each channel that have undergone noise reduction processing are weighted and averaged according to the distance from the sound source position to obtain a clean and smooth voice signal.
  • the control equipment includes: Collecting the sound information collected by all microphones in the distributed microphone array and processed by the voice processing module; Sending the collected sound information to the background data collector after processing; Connecting the data collector to the equipment and collecting data according to The sound information received by the receiver controls the device.
  • An embodiment of the present application provides an electronic device. Please refer to 5.
  • the electronic device includes a memory 601, a processor 602, and a computer program stored in the memory 601 and running on the processor 602.
  • the processor 602 When the computer program is executed, the distributed microphone array pickup method described above is realized.
  • the electronic device further includes: at least one input device 603 and at least one output device 604.
  • the aforementioned memory 601, processor 602, input device 603, and output device 604 are connected through a bus 605.
  • the input device 603 may specifically be a camera, a touch panel, a physical button or a mouse, etc.
  • the output device 604 may specifically be a display screen.
  • the memory 601 may be a high-speed random access memory (RAM, Random Access Memory) memory, or may be a non-volatile memory (non-volatile memory), such as a disk memory.
  • RAM Random Access Memory
  • non-volatile memory such as a disk memory.
  • the memory 601 is used to store a group of executable program codes, and the processor 602 is coupled with the memory 601.
  • an embodiment of the present application also provides a computer-readable storage medium.
  • the computer-readable storage medium may be the electronic device provided in the foregoing embodiments, and the computer-readable storage medium may be the aforementioned ⁇ Memory 601.
  • a computer program is stored on the computer-readable storage medium, and when the program is executed by the processor 602, the distributed microphone array pickup method described in the foregoing method embodiment is implemented.
  • the computer storage medium may also be a U disk, a mobile hard disk, a read-only memory 601 (ROM, Read-Only Memory), RAM, a magnetic disk, or an optical disk, and other media that can store program codes.
  • the disclosed apparatus and method may be implemented in other ways.
  • the device embodiments described above are merely illustrative, for example, the division of the modules is only a logical function division, and there may be other division methods in actual implementation, for example, multiple modules or components may be combined or It can be integrated into another system, or some features can be ignored or not implemented.
  • the displayed or discussed mutual coupling or direct coupling or communication connection may be indirect coupling or communication connection through some interfaces, devices or modules, and may be electrical , Mechanical or other forms.
  • modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical modules, that is, they may be located in one place, or they may be distributed to multiple networks On the module. Some or all of the modules can be selected according to actual needs to achieve the objectives of the solution of the embodiment.
  • the functional modules in the various embodiments of the present invention may be integrated into one processing module, or each module may exist alone physically, or two or more modules may be integrated into one module.
  • the above-mentioned integrated modules can be implemented in the form of hardware or software functional modules.
  • the integrated module is implemented in the form of a software function module and sold or used as an independent product, it can be stored in a computer readable storage medium.
  • the technical solution of the present invention essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium. , Including several instructions to enable a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in the various embodiments of the present invention.

Landscapes

  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Acoustics & Sound (AREA)
  • Signal Processing (AREA)
  • Circuit For Audible Band Transducer (AREA)
  • Obtaining Desirable Characteristics In Audible-Bandwidth Transducers (AREA)

Abstract

本发明公开了一种分布式麦克风阵列拾音***及方法,用于拾取声音信息来控制设备,解决了现有的语音助手拾音效果差,且设备均与麦克风拾音模块及语音助手绑定在一起而造成了资源浪费的问题,包括:分布式麦克风阵列,用于在室内收集声音信息;声源定位处理模块,用于定位说话人发出声音信息的声源位置;语音处理模块,用于根据声源位置及分布式麦克风阵列收集的声音信息进行处理;麦克风数据采集模块,用于采集分布式麦克风阵列发射的信号,并将分布式麦克风阵列发射的信号同步到一个数据采集终端;本***拾音效果好,并能通过语音信息控制多个设备,而不必为每个设备配套安装独立的语音拾取模块及语音助手,从而减少了资源的浪费。

Description

一种分布式麦克风阵列拾音***及方法 技术领域
[0001] 本发明涉及语音识别技术领域, 尤其涉及一种分布式麦克风阵列拾音***及方 法。
背景技术
[0002] 随着人工智能技术的发展, 通过人机交互的方式来控制设备的技术也越来越普 遍, 而在人机交互技术中, 通过语音控制设备的技术得到了广泛的应用, 而在 智能家居等应用场景中, 语音助手几乎成了未来发展的标准配置。
[0003] 然而, 想要设备拥有类似人类的听觉能力并不容易, 除了后端的语音识别、 语 义理解等诸多环节外, 第一步就是要让设备获取清晰的语音信号, 即让设备听 得更清楚。
[0004] 5见有的设备配备的语音助手, 安装位置比较局限只有一个点, 对于较远距离的 拾音效果较差, 而且现有的设备均与语音助手绑定在一起, 当一个房间中放置 了多台设备时, 就需要配置多个语音助手, 因此造成了资源的浪费。
发明概述
技术问题
[0005] 本发明的主要目的在于提供一种分布式麦克风阵列拾音***及方法, 旨在解决 5见有技术中的语音助手拾音效果差, 且设备均与语音助手绑定在一起而造成了 资源浪费的技术问题。
问题的解决方案
技术解决方案
[0006] 为实现上述目的, 本发明第一方面提供一种分布式麦克风阵列拾音***, 包括 : 分布式麦克风阵列, 用于在室内收集声音信息; 声源定位处理模块, 用于定 位说话人发出声音信息的声源位置; 语音处理模块, 用于根据声源位置及分布 式麦克风阵列收集的声音信息进行处理, 得到干净平滑的语音信号; 麦克风数 据采集模块, 用于采集所述分布式麦克风阵列经过语音处理模块处理后发射的 信号, 并将所述分布式麦克风阵列发射的信号同步到一个数据采集终端, 所述 数据采集终端通过分布式麦克风阵列发射的信号控制设备。
[0007] 进一步地, 所述分布式麦克风阵列包括设置在室内屋顶和四面墙壁上安装的麦 克风; 通过调整各个麦克风的位置, 使得所述墙壁上设置的麦克风和屋顶上设 置的麦克风相互连线在空间上呈现四棱锥体的形状。
[0008] 进一步地, 所述声源定位处理模块包括: 麦克风坐标确定单元, 用于设定所述 分布式麦克风阵列内其中一个参考麦克风的三维坐标, 并根据设定的所述三维 坐标设定三维坐标系及所述分布式麦克风阵列内其他麦克风的三维坐标; 时延 估计单元, 用于根据麦克风接收到的信号的互相关函数估算出各个麦克风相对 于参考麦克风的时间延迟; 声源位置坐标确定单元, 用于根据声源到达各个麦 克风之间的时间、 延迟、 声速以及各个麦克风的坐标计算声源位置的坐标。
[0009] 进一步地, 所述语音处理模块包括: 降噪单元, 用于将所述分布式麦克风阵列 收集的声音信息进行降噪, 得到干净的语音信号; 加权平均单元, 用于将各路 麦克风信号进行延时对齐, 进而对其进行加权平均, 得到干净平滑的语音信号
[0010] 进一步地, 所述降噪单元包括: 主麦克风信号设定子单元, 用于设定主麦克风 的信号由人发出的语音信号及噪声信号组成; 麦克风确定子单元, 用于根据声 源至各个麦克风的距离确定距离声源最近的麦克风为主麦克风, 其余麦克风为 副麦克风, 并确定距离声源最远的麦克风为噪声信号参考麦克风; 噪声估计子 单元, 用于使用噪声信号参考麦克风的初始无声段信号作为噪声估计, 并将所 述噪声估计作为所述分布式麦克风阵列的噪声信号; 功率谱计算子单元, 用于 根据所述主麦克风信号设定子单元及所述噪声估计子单元计算所述主麦克风信 号的功率谱; 因子引入子单元, 用于在所述功率谱计算子单元中引入过减因子 及谱下限补偿因子; 傅里叶反变换子单元, 用于通过傅里叶反变换将引入了因 子引入单元的功率谱计算子单元的信号变换为时域信号, 从而得到干净的语音 信号。
[0011] 进一步地, 所述加权平均单元包括: 时延估计子单元, 用于根据克风接收到的 信号的互相关函数估算出各个麦克风相对于参考麦克风的时间延迟; 延时对齐 子单元, 用于根据各路信号的延迟时间将所有麦克风接收的语音信号进行延时 对齐, 从而对各路已经经过降噪处理的语音信号按照其与声源位置的距离进行 加权平均, 得到干净平滑的语音信号。
[0012] 进一步地, 所述麦克风数据采集模块包括: 数据采集单元, 用于采集所述分布 式麦克风阵列内所有麦克风收集并经过语音处理模块处理的声音信息; 数据发 送单元, 用于将所述数据采集单元采集的声音信息发送至后台数据采集器; 设 备控制单元, 用于连接设备, 并根据数据采集器接收的声音信息控制设备。
[0013] 本发明第二方面提供一种分布式麦克风阵列拾音方法, 包括: 通过分布式麦克 风阵列在室内收集声音信息; 定位说话人的声源位置; 根据声源位置及分布式 麦克风阵列收集的声音信息进行处理, 得到干净平滑的语音信号; 采集分布式 麦克风阵列发射的信号, 并将所述分布式麦克风阵列中各个麦克风发射的信号 同步到一个数据采集终端, 所述数据采集终端通过所述分布式麦克风阵列发射 的信号控制设备。
[0014] 本发明第三方面提供一种电子装置, 包括: 存储器、 处理器及存储在所述存储 器上并可在所述处理器上运行的计算机程序, 其特征在于, 所述处理器执行所 述计算机程序时, 实现上述的方法。
[0015] 本发明第四方面提供一种计算机可读存储介质, 其上存储有计算机程序, 其特 征在于, 所述计算机程序被处理器执行时, 实现上述的方法。
发明的有益效果
有益效果
[0016] 本发明提供一种分布式麦克风阵列拾音***及方法, 有益效果在于: 通过分布 式结构在室内对多个麦克风进行阵列布局, 使得室内有多个位置能够进行语音 信息的采集, 从而使用户不必为了选择特定的设备进行语音交互而需要走到该 设备附近发出声音信息, 从而提升了拾音效果; 并且在将多个设备与麦克风数 据采集模块绑定, 即, 将分布式麦克风阵列发射的信号同步到一个数据采集终 端后, 使得本***能够通过语音信息控制多个设备, 而不必为每个设备安装独 立的拾音模块和语音助手, 从而减少了资源的浪费。
对附图的简要说明 附图说明
[0017] 为了更清楚地说明本发明实施例或现有技术中的技术方案, 下面将对实施例或 5见有技术描述中所需要使用的附图作简单地介绍, 显而易见地, 下面描述中的 附图仅仅是本发明的一些实施例, 对于本领域技术人员来讲, 在不付出创造性 劳动的前提下, 还可以根据这些附图获得其他的附图。
[0018] 图 1为本发明实施例分布式麦克风阵列拾音***的结构示意框图;
[0019] 图 2为本发明实施例分布式麦克风阵列拾音***的声源定位处理模块的结构示 意框图;
[0020] 图 3为本发明实施例分布式麦克风阵列拾音***的语音处理模块的结构示意框 图;
[0021] 图 4为本发明实施例分布式麦克风阵列拾音***的降噪单元的结构示意框图;
[0022] 图 5为本发明实施例电子装置的结构示意框图。
发明实施例
本发明的实施方式
[0023] 为使得本发明的发明目的、 特征、 优点能够更加的明显和易懂, 下面将结合本 发明实施例中的附图, 对本发明实施例中的技术方案进行清楚、 完整地描述, 显然, 所描述的实施例仅仅是本发明一部分实施例, 而非全部实施例。 基于本 发明中的实施例, 本领域技术人员在没有做出创造性劳动前提下所获得的所有 其他实施例, 都属于本发明保护的范围。
[0024] 请参阅图 1, 为一种分布式麦克风阵列拾音***, 包括: 分布式麦克风阵列 1, 用于在室内收集声音信息; 声源定位处理模块 2, 用于定位说话人发出声音信息 的声源位置; 语音处理模块 3 , 用于根据声源位置及分布式麦克风阵列 1收集的 声音信息进行处理, 得到干净平滑的语音信号; 麦克风数据采集模块 4, 用于采 集分布式麦克风阵列 1经过语音处理模块 3处理后发射的信号, 并将分布式麦克 风阵列 1发射的信号同步到一个数据采集终端, 数据采集终端通过分布式麦克风 阵列 1发射的信号控制设备。
[0025] 分布式麦克风阵列 1包括设置在室内屋顶和四面墙壁上安装的麦克风; 通过调 整各个麦克风的位置, 使得墙壁上设置的麦克风和屋顶上设置的麦克风相互连 线在空间上呈现四棱锥体的形状。
[0026] 请参阅图 2, 声源定位处理模块 2包括: 麦克风坐标确定单元 21, 用于设定分布 式麦克风阵列 1内其中一个参考麦克风的三维坐标, 并根据设定的三维坐标设定 三维坐标系及分布式麦克风阵列 1内其他麦克风的三维坐标; 时延估计单元 22, 用于根据麦克风接收到的信号的互相关函数估算出各个麦克风相对于参考麦克 风的时间延迟; 声源位置坐标确定单元 23 , 用于根据声源到达各个麦克风之间 的时间、 延迟、 声速以及各个麦克风的坐标计算声源位置的坐标。
[0027] 具体地, 设定矩形房间在三维空间坐标系内沿 x轴、 y轴及 z轴方向的尺寸分别 为 a、 b、 c, 房间的四面墙壁和天花板上各有一个麦克风组成一个分布式麦克风 阵列 1, 设定其中一个麦克风的坐标为
fru yu %)
, 记做第一麦克风, 其他的麦克风坐标为
其中, m=2, 3,4, 5 , 设定说话人发出声音信息的声源位
Figure imgf000007_0001
, 并设定声音信号从声源位置到达各个麦克风所需的时间为
, 设定声速为 V
, 则声源位置到各个麦克风的距离为
4 , 则可得到:
[0028]
Figure imgf000008_0001
Figure imgf000008_0002
[0029] 因此可以计算声音信号从声源位置到达分布式麦克风阵列的第 m麦克风与到达 第一麦克风的时间延迟为:
Figure imgf000008_0003
wt m
[0041] 使用公式 2减去公式 1得到公式 3, 公式 3表示如下: [0042]
(h : m视 . wi).
Figure imgf000009_0001
m: 2
[0043] 由于声源位置与麦克风的坐标分别为
Figure imgf000009_0002
m v v %
Jwt (tn J
, 则声源位置到第 m麦克风的距离可以表示为:
Figure imgf000009_0003
[0046] 同理可以将声源位置到第一麦克风的距离表示为:
Figure imgf000009_0004
[0048] 因此使用
[0050] 再将公式 4代入公式 3得到公式 5, 公式 5表示如下:
[0051]
Figure imgf000010_0001
[0052] 此时设定四个变量分别为
Mm a tn
/ m
及 i m
, 并设定四个变量的表达式如下:
Figure imgf000010_0002
[0059] 而对于公式 6中的参数
Figure imgf000011_0001
则可以通过基于广义互相关 (GCC-PHAT) 算法的时延估计 (TDOA) 方法计 算得到。 由于本申请提供的***采用了分布式麦克风阵列 1, 其中的麦克风相互 之间的间距较大, 阵列信号的空间分辨率很高, 采用 GCC-PHAT算法估计得到 的延时
7
f TO
也较为准确。 因此, 当 m=3,4,5时, 使用公式 6组成一个三元一次方程组, 并表 示如下: [0062] 请参阅图 3及图 4, 语音处理模块 3包括: 降噪单元 31, 用于将分布式麦克风阵 歹 IJ1收集的声音信息进行降噪, 得到干净的语音信号; 加权平均单元 32, 用于将 各路麦克风信号进行延时对齐, 进而对其进行加权平均, 得到干净平滑的语音 信号。
[0063] 降噪单元 31包括: 主麦克风信号设定子单元 311, 用于设定主麦克风的信号由 人发出的语音信号及噪声信号组成; 麦克风确定子单元 312, 用于根据声源到达 各个麦克风的距离确定距离声源最近的麦克风为主麦克风, 其余麦克风为副麦 克风, 并确定距离声源最远的麦克风为噪声信号参考麦克风; 噪声估计子单元 3 13 , 用于使用噪声信号参考麦克风的初始无声段信号作为噪声估计, 并将噪声 估计作为分布式麦克风阵列的噪声信号; 功率谱计算子单元 314, 用于根据主麦 克风信号设定子单元 311及噪声估计子单元 313计算各麦克风信号的功率谱; 因 子引入子单元 315 , 用于在功率谱计算子单元 314中引入过减因子及谱下限补偿 因子; 傅里叶反变换子单元 316 , 用于通过傅里叶反变换将引入了因子引入单元 的功率谱计算子单元 314的信号变换为时域信号, 得到干净的语音信号。
[0064] 在得到声源位置的坐标后, 通过计算声源位置与各个麦克风之间的距离
, 确定距离声源位置最近的麦克风为主麦克风, 其他麦克风为副麦克风, 并将 距离声源最远的麦克风作为噪声信号参考麦克风。
[0065] 具体地, 设定主麦克风信号
Figure imgf000012_0001
, 其中
X(11)
为目标语音信号,
¥(11〕
噪声信号, 设噪声信号参考麦克风的初始无声段信号为 。 由于 5个麦克风同时工作, 因此所有麦克风接收到的噪声信号是相同的。 利用 噪声信号参考麦克风的初始无声段来做为噪声估计, 因此
y(n) : : x(n) ! ¥ (ii)
, 其对应的功率谱表达式为:
Figure imgf000013_0001
[0068] 使用谱减得到语音信号功率谱的过程中可能会由于过减而导致功率谱为负值, 因此引入了过减因子
Figure imgf000013_0002
[0070] 而由于人的听觉对语音的相位不敏感, 因此通过谱减降噪得到语音信号的幅度 谱
Figure imgf000013_0003
可以跟麦克风接收到的原始信号的相位谱复合为估计语音信号的频谱, 再通过 傅里叶反变换就可以还原得到干净的语音信号
t(ll)
[0071] 加权平均单元 32包括: 时延估计子单元 321, 用于根据麦克风接收到的信号的 互相关函数估算出各个麦克风相对于参考麦克风的时间延迟; 延时对齐子单元 3 22, 用于根据各路信号的延迟时间将所有麦克风接收的语音信号进行延时对齐 , 从而对各路已经经过降噪处理的语音信号按照其与声源位置的距离进行加权 平均, 得到干净平滑的语音信号。
[0072] 经过上述的降噪处理之后, 各个麦克风通道均可得到比较干净的语音信号, 为 了增强说话人方向上的语音信号, 假设说话人在短时间内的位置不变, 使用各 个麦克风的相对时延, 将五路麦克风信号进行延时对齐, 然后再根据各自距离 说话人的距离进行加权平均, 就可以得到干净平滑的语音信号。
[0073]
Figure imgf000014_0001
[0074] 麦克风数据采集模块 4包括: 数据采集单元 41, 用于采集分布式麦克风阵列 1内 所有麦克风收集并经过语音处理模块 3处理的声音信号; 数据发送单元 42, 用于 将数据采集单元 41采集的声音信息发送至后台数据采集器; 设备控制单元 43, 用于连接设备, 并根据数据采集器接收的声音信息控制设备。
[0075] 在本实施例中, 分布式麦克风阵列 1与数据采集单元 41通过无线网络连接, 在 麦克风接收到的语音信号经过语音处理模块 3处理后得到干净平滑的语音信号, 再通过无线网络将语音信号传输至数据采集单元 41, 数据发送单元 42将数据采 集单元 41接收的语音信号传输至后台的数据采集器, 设备控制单元 43在连接了 设备后, 能够通过数据采集器接收的语音信号控制设备。
[0076] 通过将数据采集单元 41与分布式麦克风阵列 1通过无线网络连接, 不仅方便组 网, 增加了智能家居语音控制中的灵活性, 同时, 采用统一的设备控制单元 43 作为中央处理平台, 可以为各智能语音交互设备提供公用的硬件输入接口, 从 而使本***能够通过语音信息控制多个设备, 而不必为每个设备安装拾音模块 和语音助手, 减少了资源的浪费。
[0077] 本申请还提供一种分布式麦克风阵列拾音方法, 包括: 通过分布式麦克风阵列 在室内收集声音信息; 定位说话人的声源位置; 根据声源位置及分布式麦克风 阵列收集的声音信息进行处理, 得到干净平滑的语音信号; 采集分布式麦克风 阵列发射的信号, 并将分布式麦克风阵列麦克风发射的信号同步到一个数据采 集终端, 数据采集终端通过分布式麦克风阵列发射的信号控制设备。 [0078] 分布式麦克风阵列包括设置在室内屋顶和四面墙壁上的麦克风; 通过调整各个 麦克风的位置, 使得墙壁上安装的麦克风和屋顶上安装的麦克风相互连线在空 间上呈现四棱锥体的形状。
[0079] 定位说话人的声源位置包括: 设定分布式麦克风阵列内其中一个参考麦克风的 三维坐标, 并根据设定的三维坐标设定三维坐标系及分布式麦克风阵列内其他 麦克风的三维坐标; 根据麦克风接收到的信号的互相关函数估算出各个麦克风 相对于参考麦克风的时间延迟; 根据声源到达各个麦克风之间的时间、 延迟、 声速以及各个麦克风的坐标计算声源位置的坐标。
[0080] 根据声源位置及分布式麦克风阵列收集的声音信息进行处理, 得到干净平滑的 语音信号包括: 将分布式麦克风阵列收集的声音信息进行降噪, 得到干净的语 音信号; 对分布式麦克风阵列中的各个麦克风的语音信号进行延时对齐和加权 平均, 得到干净平滑的语音信号。
[0081] 将分布式麦克风阵列收集的声音信息进行降噪, 得到干净的语音信号包括: 设 定主麦克风的信号由人发出的语音信号及噪声信号组成; 根据声源到达各个麦 克风的距离确定距离声源最近的麦克风为主麦克风, 其余麦克风为副麦克风, 并确定距离声源最远的麦克风为噪声信号参考麦克风; 使用噪声信号参考麦克 风的初始无声段信号作为噪声估计, 并将噪声估计作为分布式麦克风阵列的噪 声信号; 根据主麦克风信号设定子单元及噪声估计子单元计算主麦克风信号的 功率谱; 在计算主麦克风信号的功率谱过程中引入过减因子及谱下限补偿因子 ; 通过傅里叶反变换将引入了因子引入单元的功率谱计算子单元的信号变换为 时域信号, 从而得到干净的语音信号。
[0082] 对分布式麦克风阵列中的各个麦克风的语音信号进行处理, 得到干净平滑的语 音信号包括: 根据克风接收到的信号的互相关函数估算出各个麦克风相对于参 考麦克风的时间延迟; 根据各路信号的延迟时间将所有麦克风接收的语音信号 进行延时对齐, 从而对各路已经经过降噪处理的语音信号按照其与声源位置的 距离进行加权平均, 得到干净平滑的语音信号。
[0083] 采集分布式麦克风阵列发射的信号, 并将分布式麦克风阵列麦克风发射的信号 同步到一个数据采集终端, 数据采集终端通过分布式麦克风阵列发射的信号控 制设备包括: 采集分布式麦克风阵列内所有麦克风收集并经过语音处理模块处 理的声音信息; 将采集的声音信息经过处理后发送至后台数据采集器; 将数据 采集器与设备连接, 并根据数据采集器接收的声音信息控制设备。
[0084] 本申请实施例提供一种电子装置, 请参阅 5 , 该电子装置包括: 存储器 601、 处 理器 602及存储在存储器 601上并可在处理器 602上运行的计算机程序, 处理器 60 2执行该计算机程序时, 实现前述中描述的分布式麦克风阵列拾音方法。
[0085] 进一步的, 该电子装置还包括: 至少一个输入设备 603以及至少一个输出设备 6 04。
[0086] 上述存储器 601、 处理器 602、 输入设备 603以及输出设备 604, 通过总线 605连 接。
[0087] 其中, 输入设备 603具体可为摄像头、 触控面板、 物理按键或者鼠标等等。 输 出设备 604具体可为显示屏。
[0088] 存储器 601可以是高速随机存取记忆体 (RAM, Random Access Memory) 存储 器, 也可为非不稳定的存储器 (non-volatile memory) , 例如磁盘存储器。 存储 器 601用于存储一组可执行程序代码, 处理器 602与存储器 601耦合。
[0089] 进一步的, 本申请实施例还提供了一种计算机可读存储介质, 该计算机可读存 储介质可以是设置于上述各实施例中的电子装置中, 该计算机可读存储介质可 以是前述的存储器 601。 该计算机可读存储介质上存储有计算机程序, 该程序被 处理器 602执行时实现前述方法实施例中描述的分布式麦克风阵列拾音方法。
[0090] 进一步的, 该计算机可存储介质还可以是 U盘、 移动硬盘、 只读存储器 601 (R OM, Read-Only Memory) 、 RAM、 磁碟或者光盘等各种可以存储程序代码的介 质。
[0091] 在本申请所提供的几个实施例中, 应该理解到, 所揭露的装置和方法, 可以通 过其它的方式实现。 例如, 以上所描述的装置实施例仅仅是示意性的, 例如, 所述模块的划分, 仅仅为一种逻辑功能划分, 实际实现时可以有另外的划分方 式, 例如多个模块或组件可以结合或者可以集成到另一个***, 或一些特征可 以忽略, 或不执行。 另一点, 所显示或讨论的相互之间的耦合或直接耦合或通 信连接可以是通过一些接口, 装置或模块的间接耦合或通信连接, 可以是电性 , 机械或其它的形式。
[0092] 所述作为分离部件说明的模块可以是或者也可以不是物理上分开的, 作为模块 显示的部件可以是或者也可以不是物理模块, 即可以位于一个地方, 或者也可 以分布到多个网络模块上。 可以根据实际的需要选择其中的部分或者全部模块 来实现本实施例方案的目的。
[0093] 另外, 在本发明各个实施例中的各功能模块可以集成在一个处理模块中, 也可 以是各个模块单独物理存在, 也可以两个或两个以上模块集成在一个模块中。 上述集成的模块既可以采用硬件的形式实现, 也可以采用软件功能模块的形式 实现。
[0094] 所述集成的模块如果以软件功能模块的形式实现并作为独立的产品销售或使用 时, 可以存储在一个计算机可读取存储介质中。 基于这样的理解, 本发明的技 术方案本质上或者说对现有技术做出贡献的部分或者该技术方案的全部或部分 可以以软件产品的形式体现出来, 该计算机软件产品存储在一个存储介质中, 包括若干指令用以使得一台计算机设备 (可以是个人计算机, 服务器, 或者网 络设备等) 执行本发明各个实施例所述方法的全部或部分步骤。
[0095] 需要说明的是, 对于前述的各方法实施例, 为了简便描述, 故将其都表述为一 系列的动作组合, 但是本领域技术人员应该知悉, 本发明并不受所描述的动作 顺序的限制, 因为依据本发明, 某些步骤可以采用其它顺序或者同时进行。 其 次, 本领域技术人员也应该知悉, 说明书中所描述的实施例均属于优选实施例 , 所涉及的动作和模块并不一定都是本发明所必须的。
[0096] 在上述实施例中, 对各个实施例的描述都各有侧重, 某个实施例中没有详述的 部分, 可以参见其它实施例的相关描述。
[0097] 以上为对本发明所提供的一种分布式麦克风阵列拾音***及方法的描述, 对于 本领域的技术人员, 依据本发明实施例的思想, 在具体实施方式及应用范围上 均会有改变之处, 综上, 本说明书内容不应理解为对本发明的限制。

Claims

权利要求书
[权利要求 1] 一种分布式麦克风阵列拾音***, 其特征在于, 包括:
分布式麦克风阵列, 用于在室内收集声音信息; 声源定位处理模块, 用于定位说话人发出声音信息的声源位置; 语音处理模块, 用于根据声源位置及分布式麦克风阵列收集的声音信 息进行处理, 得到干净平滑的语音信号;
麦克风数据采集模块, 用于采集所述分布式麦克风阵列经过语音处理 模块处理后发射的信号, 并将所述分布式麦克风阵列发射的信号同步 到一个数据采集终端, 所述数据采集终端通过分布式麦克风阵列发射 的信号控制设备。
[权利要求 2] 根据权利要求 1所述的分布式麦克风阵列拾音***, 其特征在于, 所述分布式麦克风阵列包括设置在室内屋顶和四面墙壁上安装的麦克 风;
通过调整各个麦克风的位置, 使得所述墙壁上设置的麦克风和屋顶上 设置的麦克风相互连线在空间上呈现四棱锥体的形状。
[权利要求 3] 根据权利要求 1所述的分布式麦克风阵列拾音***, 其特征在于, 所述声源定位处理模块包括:
麦克风坐标确定单元, 用于设定所述分布式麦克风阵列内其中一个参 考麦克风的三维坐标, 并根据设定的所述三维坐标设定三维坐标系及 所述分布式麦克风阵列内其他麦克风的三维坐标; 时延估计单元, 用于根据麦克风接收到的信号的互相关函数估算出各 个麦克风相对于参考麦克风的时间延迟;
声源位置坐标确定单元, 用于根据声源到达各个麦克风之间的时间、 延迟、 声速以及各个麦克风的坐标计算声源位置的坐标。
[权利要求 4] 根据权利要求 1所述的分布式麦克风阵列拾音***, 其特征在于, 所述语音处理模块包括:
降噪单元, 用于将所述分布式麦克风阵列收集的声音信息进行降噪, 得到干净的语音信号; 加权平均单元, 用于将各路麦克风信号进行延时对齐, 进而对其进行 加权平均, 得到干净平滑的语音信号。
[权利要求 5] 根据权利要求 4所述的分布式麦克风阵列拾音***, 其特征在于, 所述降噪单元包括:
主麦克风信号设定子单元, 用于设定主麦克风的信号由人发出的语音 信号及噪声信号组成;
麦克风确定子单元, 用于根据声源到达各个麦克风的距离确定距离声 源最近的麦克风为主麦克风, 其余麦克风为副麦克风, 并确定距离声 源最远的麦克风为噪声信号参考麦克风;
噪声估计子单元, 用于使用噪声信号参考麦克风的初始无声段信号作 为噪声估计, 并将所述噪声估计作为所述分布式麦克风阵列的噪声信 号;
功率谱计算子单元, 用于根据所述主麦克风信号设定子单元及所述噪 声估计子单元计算所述各麦克风信号的功率谱; 因子引入子单元, 用于在所述功率谱计算子单元中引入过减因子及谱 下限补偿因子;
傅里叶反变换子单元, 用于通过傅里叶反变换将弓 I入了因子引入单元 的功率谱计算子单元的信号变换为时域信号, 从而得到干净的语音信 号。
[权利要求 6] 根据权利要求 5所述的分布式麦克风阵列拾音***, 其特征在于, 所述加权平均单元包括:
时延估计子单元, 用于根据克风接收到的信号的互相关函数估算出各 个麦克风相对于参考麦克风的时间延迟;
延时对齐子单元, 用于根据各路信号的延迟时间将所有麦克风接收的 语音信号进行延时对齐, 从而对各路已经经过降噪处理的语音信号按 照其与声源位置的距离进行加权平均, 得到干净平滑的语音信号。
[权利要求 7] 根据权利要求 1所述的分布式麦克风阵列拾音***, 其特征在于, 所述麦克风数据采集模块包括: 数据采集单元, 用于采集所述分布式麦克风阵列内所有麦克风收集并 经过语音处理模块处理的声音信号;
数据发送单元, 用于将所述数据采集单元采集的声音信息发送至后台 数据采集器;
设备控制单元, 用于连接设备, 并根据数据采集器接收的声音信息控 制设备。
[权利要求 8] 一种分布式麦克风阵列拾音方法, 其特征在于, 包括:
通过分布式麦克风阵列在室内收集声音信息;
定位说话人的声源位置;
根据声源位置及分布式麦克风阵列收集的声音信息进行处理, 得到干 净平滑的语音信号;
采集分布式麦克风阵列发射的信号, 并将所述分布式麦克风阵列中各 个麦克风发射的信号同步到一个数据采集终端, 所述数据采集终端通 过所述分布式麦克风阵列发射的信号控制设备。
[权利要求 9] 一种电子装置, 包括: 存储器、 处理器及存储在所述存储器上并可在 所述处理器上运行的计算机程序, 其特征在于, 所述处理器执行所述 计算机程序时, 实现权利要求 8中的所述方法。
[权利要求 10] 一种计算机可读存储介质, 其上存储有计算机程序, 其特征在于, 所 述计算机程序被处理器执行时, 实现权利要求 8中的所述方法。
PCT/CN2019/086768 2019-01-25 2019-05-14 一种分布式麦克风阵列拾音***及方法 WO2020151133A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201910073427.7A CN109788382A (zh) 2019-01-25 2019-01-25 一种分布式麦克风阵列拾音***及方法
CN201910073427.7 2019-01-25

Publications (1)

Publication Number Publication Date
WO2020151133A1 true WO2020151133A1 (zh) 2020-07-30

Family

ID=66501290

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/086768 WO2020151133A1 (zh) 2019-01-25 2019-05-14 一种分布式麦克风阵列拾音***及方法

Country Status (2)

Country Link
CN (1) CN109788382A (zh)
WO (1) WO2020151133A1 (zh)

Families Citing this family (25)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110234043B (zh) * 2019-05-31 2020-08-25 歌尔科技有限公司 基于麦克风阵列的声音信号处理方法、装置及设备
CN110197671A (zh) * 2019-06-17 2019-09-03 深圳壹秘科技有限公司 定向拾音方法、录音设备和存储介质
CN110418242B (zh) * 2019-07-30 2021-02-05 西安声必捷信息科技有限公司 声源定向方法、装置及***
CN110519676B (zh) * 2019-08-22 2021-04-09 云知声智能科技股份有限公司 一种去中心化的分布式麦克风拾音的方法
CN111161751A (zh) * 2019-12-25 2020-05-15 声耕智能科技(西安)研究院有限公司 复杂场景下的分布式麦克风拾音***及方法
CN113141285B (zh) * 2020-01-19 2022-04-29 海信集团有限公司 一种沉浸式语音交互方法及***
CN111246324A (zh) * 2020-02-24 2020-06-05 江苏中协智能科技有限公司 定向拾音吊麦用辅助装置及使用者身份确认、追踪方法
CN111544030B (zh) * 2020-05-20 2023-06-20 京东方科技集团股份有限公司 一种听诊器、诊断装置及诊断方法
CN111768797A (zh) * 2020-09-01 2020-10-13 深圳市声扬科技有限公司 语音增强处理方法、装置、计算机设备和存储介质
CN112565531B (zh) * 2020-12-12 2021-08-13 深圳波导智慧科技有限公司 一种应用于多人语音会议的录音方法和设备
CN112769599B (zh) * 2020-12-28 2022-11-22 深信服科技股份有限公司 一种资源自动接入方法、***及可读存储介质
CN112735462B (zh) * 2020-12-30 2024-05-31 科大讯飞股份有限公司 分布式麦克风阵列的降噪方法和语音交互方法
CN112887875B (zh) * 2021-01-22 2022-10-18 平安科技(深圳)有限公司 会议***语音数据采集方法、装置、电子设备及存储介质
CN113068101B (zh) * 2021-03-16 2023-06-09 Oppo广东移动通信有限公司 指环阵列拾音控制方法、装置、存储介质及指环阵列
CN113516981A (zh) * 2021-03-31 2021-10-19 深圳鱼亮科技有限公司 一种基于无线分布式麦克风阵列的语音唤醒决策***
CN113092965A (zh) * 2021-04-09 2021-07-09 华北电力大学(保定) 一种基于麦克风阵列的换流阀设备缺陷局部放电检测装置
CN113311390A (zh) * 2021-04-25 2021-08-27 普联国际有限公司 基于分布式麦阵的声源定位方法、装置、设备及存储介质
CN113301461A (zh) * 2021-05-14 2021-08-24 深圳波洛斯科技有限公司 一种基于环境音消除的主动降噪方法及其***
CN113608167B (zh) * 2021-10-09 2022-02-08 阿里巴巴达摩院(杭州)科技有限公司 声源定位方法、装置及设备
CN114268883A (zh) * 2021-11-29 2022-04-01 苏州君林智能科技有限公司 一种选择麦克风布放位置的方法与***
CN114205731B (zh) * 2021-12-08 2023-12-26 随锐科技集团股份有限公司 发言人区域检测方法、装置、电子设备和存储介质
CN114325586B (zh) * 2021-12-31 2024-05-14 智己汽车科技有限公司 一种车载多媒体声音定位***
CN115002607B (zh) * 2022-05-25 2023-12-22 歌尔股份有限公司 音源位置确定方法、设备及存储介质
CN114974200B (zh) * 2022-07-13 2023-02-03 恩平市菲玛特电子科技有限公司 基于物联网的麦克风语音交互管理***及方法
CN114923136A (zh) * 2022-07-15 2022-08-19 安徽云磬科技产业发展有限公司 一种多阵列管道泄漏定位方法与装置

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20100008516A1 (en) * 2008-07-11 2010-01-14 International Business Machines Corporation Method and system for position detection of a sound source
CN204390737U (zh) * 2014-07-29 2015-06-10 科大讯飞股份有限公司 一种家庭语音处理***
CN105866741A (zh) * 2016-06-23 2016-08-17 合肥联宝信息技术有限公司 基于声源定位的家居控制装置及方法
CN106507244A (zh) * 2016-12-23 2017-03-15 深圳先进技术研究院 一种中控***
CN107682240A (zh) * 2017-09-27 2018-02-09 四川长虹电器股份有限公司 一种用于智慧家居的分布式语音交互***

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN206353839U (zh) * 2016-12-13 2017-07-25 深圳创维-Rgb电子有限公司 一种电视语音控制***
CN107123421A (zh) * 2017-04-11 2017-09-01 广东美的制冷设备有限公司 语音控制方法、装置及家电设备
CN107290723B (zh) * 2017-06-22 2019-11-05 北京地平线信息技术有限公司 声源定位方法、装置和电子设备
CN108680902A (zh) * 2018-06-27 2018-10-19 相舆科技(上海)有限公司 一种基于多麦克风阵列的声源定位***

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20100008516A1 (en) * 2008-07-11 2010-01-14 International Business Machines Corporation Method and system for position detection of a sound source
CN204390737U (zh) * 2014-07-29 2015-06-10 科大讯飞股份有限公司 一种家庭语音处理***
CN105866741A (zh) * 2016-06-23 2016-08-17 合肥联宝信息技术有限公司 基于声源定位的家居控制装置及方法
CN106507244A (zh) * 2016-12-23 2017-03-15 深圳先进技术研究院 一种中控***
CN107682240A (zh) * 2017-09-27 2018-02-09 四川长虹电器股份有限公司 一种用于智慧家居的分布式语音交互***

Also Published As

Publication number Publication date
CN109788382A (zh) 2019-05-21

Similar Documents

Publication Publication Date Title
WO2020151133A1 (zh) 一种分布式麦克风阵列拾音***及方法
WO2021037129A1 (zh) 一种声音采集方法及装置
Zhang et al. Why does PHAT work well in lownoise, reverberative environments?
WO2017185663A1 (zh) 一种增加混响的方法与装置
CN111239687B (zh) 一种基于深度神经网络的声源定位方法及***
JP2019518985A (ja) 分散したマイクロホンからの音声の処理
WO2016014254A1 (en) System and method for determining audio context in augmented-reality applications
WO2020048431A1 (zh) 一种语音处理方法、电子设备和显示设备
US20180350357A1 (en) System and methods for audio pattern recognition
CN110072177B (zh) 空间划分信息获取方法、装置及存储介质
CN103458220A (zh) 一种图像处理方法以及电子设备
WO2016119388A1 (zh) 一种基于语音信号构造聚焦协方差矩阵的方法及装置
Hao et al. Spectral flux-based convolutional neural network architecture for speech source localization and its real-time implementation
CN114830681A (zh) 用于减少环境噪声补偿***中的误差的方法
CN106339081A (zh) 一种基于商用设备的免携带设备手掌定位人机交互方法
CN103873978A (zh) 信息处理的方法及电子设备
CN111540365B (zh) 语音信号确定方法、装置、服务器及存储介质
CN112363112A (zh) 一种基于线性麦克风阵列的声源定位方法及装置
CN111932619A (zh) 结合图像识别和语音定位的麦克风跟踪***及方法
Aprea et al. Acoustic reconstruction of the geometry of an environment through acquisition of a controlled emission
WO2022170949A1 (zh) 设备间的相对位置检测方法及装置
CN107277690B (zh) 声音处理方法、装置和电子设备
US10152985B2 (en) Method for recording in video chat, and terminal
CN113035174A (zh) 语音识别处理方法、装置、设备及***
WO2023088156A1 (zh) 一种声速矫正方法以及装置

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19911749

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 22/11/2021)

122 Ep: pct application non-entry in european phase

Ref document number: 19911749

Country of ref document: EP

Kind code of ref document: A1