WO2020238203A1 - 降噪方法、降噪装置及可实现降噪的设备 - Google Patents

降噪方法、降噪装置及可实现降噪的设备 Download PDF

Info

Publication number
WO2020238203A1
WO2020238203A1 PCT/CN2019/129930 CN2019129930W WO2020238203A1 WO 2020238203 A1 WO2020238203 A1 WO 2020238203A1 CN 2019129930 W CN2019129930 W CN 2019129930W WO 2020238203 A1 WO2020238203 A1 WO 2020238203A1
Authority
WO
WIPO (PCT)
Prior art keywords
signal
sound
noise reduction
noise
target
Prior art date
Application number
PCT/CN2019/129930
Other languages
English (en)
French (fr)
Inventor
陈孝良
冯大航
苏少炜
常乐
Original Assignee
北京声智科技有限公司
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 北京声智科技有限公司 filed Critical 北京声智科技有限公司
Publication of WO2020238203A1 publication Critical patent/WO2020238203A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G10—MUSICAL INSTRUMENTS; ACOUSTICS
    • G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
    • G10L15/00—Speech recognition
    • G10L15/20—Speech recognition techniques specially adapted for robustness in adverse environments, e.g. in noise, of stress induced speech
    • G—PHYSICS
    • G10—MUSICAL INSTRUMENTS; ACOUSTICS
    • G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
    • G10L15/00—Speech recognition
    • G10L15/22—Procedures used during a speech recognition process, e.g. man-machine dialogue

Definitions

  • the embodiments of the present disclosure relate to a noise reduction method, a noise reduction device, and a device that can realize noise reduction.
  • At least one embodiment of the present disclosure provides a noise reduction method, including: acquiring a sound input signal received by a target device; using the external noise signal of the target device to perform noise reduction processing on the sound input signal to obtain an external target sound Signal; using the internal noise signal of the target device to perform noise reduction processing on the external target sound signal to obtain the target sound signal.
  • the acquiring the sound input signal received by the target device includes: acquiring the initial sound signal input from the outside of the target device to the target device; Perform noise reduction processing to obtain the voice input signal.
  • the acquiring the sound input signal received by the target device includes: acquiring the initial sound signal input from the outside of the target device to the target device; and combining the initial sound signal As the voice input signal.
  • the noise reduction processing is performed by using a fixed beam form or an adaptive filter.
  • the internal noise signal is obtained through a first sound acquisition module provided inside the target device; the first sound acquisition module is insulated from the outside of the target device.
  • a casing is sheathed outside the target device, and the casing is soundproof.
  • At least one embodiment of the present disclosure further provides a noise reduction method, including: acquiring a sound input signal received by a target device; using the internal noise signal of the target device to perform noise reduction processing on the sound input signal to obtain a first Sound signal; using the external noise signal of the target device to perform noise reduction processing on the first sound signal to obtain the target sound signal.
  • At least one embodiment of the present disclosure provides a noise reduction method, including: acquiring a sound input signal received by a target device; using the external noise signal and internal noise signal of the target device to input the sound in the same noise reduction operation The signal undergoes noise reduction processing to obtain the target sound signal.
  • At least one embodiment of the present disclosure provides a noise reduction method, including: acquiring a sound input signal received by a target device; using the external noise signal of the target device as the only reference signal, and performing noise reduction processing on the sound input signal , Get the target sound signal.
  • At least one embodiment of the present disclosure provides a noise reduction method, including: acquiring a sound input signal received by a target device; using the internal noise signal of the target device as the only reference signal, and performing noise reduction processing on the sound input signal , Get the target sound signal.
  • At least one embodiment of the present disclosure provides a noise reduction device, including: a signal acquisition unit, a first noise reduction processing unit, and a second noise reduction processing unit; wherein the signal acquisition unit is configured to acquire a sound input received by a target device Signal; the first noise reduction processing unit is configured to use the external noise signal of the target device to perform noise reduction processing on the sound input signal to obtain an external target sound signal; the second noise reduction processing unit is configured to , Using the internal noise signal of the target device to perform noise reduction processing on the external target sound signal to obtain the target sound signal.
  • At least one embodiment of the present disclosure provides a noise reduction device, including: a processor; a memory, including one or more computer program modules; wherein the one or more computer program modules are stored in the memory and configured To be executed by the processor, the one or more computer program modules include instructions for implementing the noise reduction method described in any embodiment of the present disclosure.
  • At least one embodiment of the present disclosure provides a device that can achieve noise reduction, including: a first sound acquisition module, a second sound acquisition module, a housing, and a noise processing module; wherein the first sound acquisition module is provided in the Inside the housing, the first sound acquisition module is soundproofed from the outside of the device; the first sound acquisition module is configured to acquire an internal noise signal inside the device, and send the internal noise signal to the Noise processing module; the second sound acquisition module is configured to acquire a sound input signal external to the device and send the sound input signal to the noise processing module; the second sound acquisition module is also configured to acquire The external noise signal external to the device and send the external noise signal to the noise processing module; the noise processing module is configured to use the external noise signal to perform noise reduction processing on the sound input signal to obtain the external For the target sound signal, the internal noise signal is then used to perform noise reduction processing on the external target sound signal to obtain the target sound signal.
  • the first sound acquisition module is provided in the Inside the housing, the first sound acquisition module is soundproofed from the outside of the device
  • the housing is soundproof.
  • the second sound acquisition module is further configured to receive an initial sound signal input externally from the device, and perform noise reduction processing on the initial sound signal to obtain the sound input signal.
  • the noise processing module uses a fixed beam form or an adaptive filter to perform noise reduction processing.
  • FIG. 1 is a schematic flowchart of a noise reduction method provided by at least one embodiment of the present disclosure
  • FIG. 2 is a schematic block diagram of a noise reduction device provided by at least one embodiment of the present disclosure
  • FIG. 3 is a schematic block diagram of another noise reduction device provided by at least one embodiment of the present disclosure.
  • Fig. 4 is a schematic block diagram of a device capable of noise reduction provided by at least one embodiment of the present disclosure.
  • At least one (item) refers to one or more, and “multiple” refers to two or more.
  • “And/or” is used to describe the association relationship of associated objects, indicating that there can be three types of relationships, for example, “A and/or B” can mean: only A, only B, and both A and B , Where A and B can be singular or plural.
  • the character “/” generally indicates that the associated objects are in an “or” relationship.
  • the following at least one item (a)” or similar expressions refers to any combination of these items, including any combination of a single item (a) or plural items (a).
  • At least one (a) of a, b or c can mean: a, b, c, "a and b", “a and c", “b and c", or "a and b and c" ", where a, b, and c can be single or multiple.
  • a smart device When a smart device implements voice recognition, it generally performs noise reduction processing on the input voice to ensure the accuracy of voice recognition.
  • the inventors of the present disclosure have discovered in research that when the internal noise of the smart device is large, the internal noise of the device will greatly interfere with the voice recognition.
  • the existing smart device still has the problem of poor noise reduction effect and cannot guarantee the voice The accuracy of recognition leads to low intelligence of the equipment.
  • the embodiments of the present disclosure provide a noise reduction method, a noise reduction device, and a device that can achieve noise reduction, which can not only eliminate the noise outside the device, but also use the second sound acquisition module set inside the device to obtain the noise inside the device.
  • Internal noise signal to eliminate the noise inside the device improve the noise elimination effect and the accuracy of voice recognition, and ensure the high intelligence of the device.
  • the sound input signal received by the target device is first acquired, and then the pre-obtained external noise signal of the target device is used to perform noise reduction processing on the sound input signal to obtain the external target sound signal. Then, the internal noise signal of the target device is used to perform noise reduction processing on the external target sound signal to obtain the target sound signal.
  • the noise reduction method provided by at least one embodiment of the present disclosure can not only avoid the influence of the external noise of the device on the accuracy of speech recognition, but also avoid the influence of the internal noise of the device on the accuracy of the speech recognition of the device, and improve the effect of noise reduction processing and the performance of the device. Intelligence.
  • FIG. 1 is a schematic flowchart of a noise reduction method provided by at least one embodiment of the present disclosure.
  • Step S101 Acquire the sound input signal received by the target device.
  • the target device can be any device that can perform voice recording, such as recording devices, devices with voice recognition, and devices that can be voice controlled, including but not limited to voice recorders, camcorders, and smart phones.
  • voice recording devices such as recording devices, devices with voice recognition, and devices that can be voice controlled, including but not limited to voice recorders, camcorders, and smart phones.
  • Smart speakers smart home equipment (such as sweeping robots, air purifiers, air conditioners, fans, etc.), smart translators, etc., are not limited here.
  • the received sound input signal includes not only the external noise signal, but also the target sound signal that is desired to be recognized (such as the voice command entered by the user).
  • the method of obtaining the sound input signal is not limited.
  • the sound input signal may be obtained by using a second sound acquisition module provided on or outside the target device.
  • the second sound acquisition module may include one or more radio devices (such as microphones of any structure).
  • the position and arrangement of the radio equipment included in the second sound acquisition module on the target equipment can be specifically set according to actual conditions and needs.
  • the second sound acquisition module may be an external microphone, or a microphone set on the casing of the target device, and the second sound acquisition module may also include a plurality of microphones arranged in a surround manner on the casing of the target device.
  • the second sound acquisition module can also be any other applicable setting method, which will not be listed here.
  • step S101 may include the following operations:
  • Step S1011 Obtain the initial sound signal of the external input target device of the target device
  • Step S1012 Perform noise reduction processing on the initial sound signal to obtain a sound input signal.
  • the initial sound signal may be the original sound signal directly received by the target device.
  • any noise reduction processing algorithm can be used to reduce the noise of the initial sound signal.
  • the noise reduction processing can be performed in the form of a fixed beam.
  • the so-called fixed beam forming means that the weight of the filter is fixed during the beam forming process.
  • step S101 may include the following operations:
  • Step S1013 Obtain the initial sound signal of the external input target device of the target device
  • Step S1014 Use the initial sound signal as a sound input signal.
  • the initial sound signal may be the original sound signal directly received by the target device, and the initial sound signal is directly used as the above-mentioned sound input signal, and the noise reduction operation on the initial sound signal is omitted.
  • the processing efficiency can be effectively improved and the amount of calculations can be reduced.
  • Step S102 Use the external noise signal of the target device to perform noise reduction processing on the sound input signal to obtain the external target sound signal.
  • the external noise signal may be an interference signal external to the target device, such as environmental noise.
  • the external noise signal is used as a reference signal for eliminating the external noise of the target device for noise reduction processing, which can eliminate the influence of the external noise of the target device on the accuracy of voice recording or recognition.
  • the external noise signal can also be obtained by the second sound acquisition module. For example, when the user does not record a voice into the target device, the sound signal received by the second sound acquisition module may be used as an external noise signal.
  • any noise reduction algorithm can be used to perform noise reduction processing, such as fixed beam or adaptive, etc., which is not limited here.
  • least mean square (LMS) adaptive filter LMS
  • normalized least mean square (NLMS) adaptive filter NLMS
  • RLS recursive least mean square
  • the principle of the LMS adaptive filter is: use the filter parameters obtained at the previous moment to automatically adjust the current filter parameters so that the input signal sequence x(n) and the expected output signal d(n) are close to adapt to the signal and Unknown or randomly changing statistical characteristics of noise to achieve optimal filtering.
  • the error signal e(n) with the input signal sequence x(n) and the expected output signal d(n) is as follows:
  • ⁇ i is the weight coefficient, n ⁇ [1, M].
  • ⁇ k (n) ⁇ k (n-1)+ ⁇ e(n)x(n) (2)
  • ⁇ is the convergence factor
  • the size of ⁇ determines the stability and convergence speed of the system.
  • the principle of other adaptive filters is similar to that of LMS. The difference is only that the standard for approximating the input signal sequence x(n) and the expected output signal d(n) is different, which will not be repeated here.
  • Step S103 Use the internal noise signal of the target device to perform noise reduction processing on the external target sound signal to obtain the target sound signal.
  • the internal noise signal may be an interference signal inside the target device, such as operating noise of the target device.
  • the internal noise signal is used as a reference signal to eliminate the internal noise of the target device for noise reduction, which can eliminate the influence of the internal noise of the target device (such as motor running noise, etc.) on the accuracy of voice input or recognition. Since the external noise signal is first used as the reference signal for noise reduction to reduce noise, the influence of external noise is reduced, and then the internal noise signal is used as the reference signal for noise reduction to perform secondary noise reduction, which reduces the influence of internal noise, thereby improving The effect and accuracy of noise reduction are improved, and the intelligence of the target device is ensured.
  • the internal noise signal can be obtained by any kind of radio equipment.
  • the internal noise signal may be obtained through a first sound acquisition module provided inside the target device.
  • the first sound acquisition module is insulated from the outside of the target device.
  • the outer casing of the target device can be set as a soundproof casing, or a soundproof mechanism for the first sound acquisition module may be added to the inside of the casing of the target device, so that the first sound acquisition module only receives The sound signal inside the target device (that is, the internal noise signal).
  • the first sound acquisition module may include at least one radio device (such as a microphone of any structure, etc.), and the position and arrangement of the at least one radio device inside the target device housing can be specifically set according to actual conditions, for example
  • the first sound acquisition module may include a plurality of microphones arranged in a surround arrangement inside the target device, which are not limited here, and will not be listed one by one.
  • any noise reduction algorithm can be used to perform noise reduction processing on the external target sound signal, such as fixed beam or adaptive, which is not limited here.
  • the sound input signal received by the target device is first acquired, and then the pre-obtained external noise signal of the target device is used to perform noise reduction processing on the sound input signal to obtain the external target sound signal. Then, the internal noise signal of the target device is used to perform noise reduction processing on the external target sound signal to obtain the target sound signal.
  • the noise reduction method provided by at least one embodiment of the present disclosure can not only avoid the influence of the external noise of the device on the accuracy of speech recognition, but also avoid the influence of the internal noise of the device on the accuracy of the speech recognition of the device, and improve the effect of noise reduction processing and the performance of the device. Intelligence.
  • At least one embodiment of the present disclosure also provides a noise reduction method, which includes the following operations.
  • Step S11 Acquire the sound input signal received by the target device
  • Step S12 Use the internal noise signal of the target device to perform noise reduction processing on the sound input signal to obtain the first sound signal;
  • Step S13 Use the external noise signal of the target device to perform noise reduction processing on the first sound signal to obtain the target sound signal.
  • the noise reduction processing in step S12 uses the internal noise signal as the reference signal
  • the noise reduction processing in step S13 uses the external noise signal as the reference signal.
  • the processing sequence of this embodiment is the same as that shown in FIG.
  • the processing sequence in the noise reduction method is different. In this way, when the internal noise is large, the internal noise signal can be used as a reference signal for noise reduction processing, thereby improving the accuracy of the final speech recognition result.
  • At least one embodiment of the present disclosure also provides a noise reduction method, which includes the following operations.
  • Step S21 Acquire the sound input signal received by the target device
  • Step S22 Use the external noise signal and internal noise signal of the target device to perform noise reduction processing on the sound input signal in the same noise reduction operation to obtain the target sound signal.
  • the noise reduction processing in step S22 uses the external noise signal and the internal noise signal as reference signals at the same time, that is, the two reference signals are used in the same noise reduction operation to realize the noise reduction processing.
  • This is different from the sequential step-by-step processing in the noise reduction method shown in Figure 1.
  • the logic of the processing algorithm can be simplified.
  • the specific method of noise reduction processing please refer to the previous content, which will not be repeated here.
  • At least one embodiment of the present disclosure also provides a noise reduction method, which includes the following operations.
  • Step S31 Acquire the sound input signal received by the target device
  • Step S32 Use the external noise signal of the target device as the only reference signal, and perform noise reduction processing on the sound input signal to obtain the target sound signal.
  • the operation can be simplified when the internal noise is small, the processing efficiency is improved, and the amount of calculation is reduced.
  • the structure of the device can be simplified.
  • At least one embodiment of the present disclosure also provides a noise reduction method, which includes the following operations.
  • Step S41 Acquire the sound input signal received by the target device
  • Step S42 Use the internal noise signal of the target device as the only reference signal, and perform noise reduction processing on the sound input signal to obtain the target sound signal.
  • the operation can be simplified when the internal noise is large and the external noise is small, the processing efficiency is improved, and the amount of calculation is reduced.
  • the specific method of noise reduction processing please refer to the previous content, which will not be repeated here.
  • At least one embodiment of the present disclosure also provides a noise reduction device.
  • FIG. 2 is a schematic block diagram of a noise reduction device provided by at least one embodiment of the present disclosure.
  • a noise reduction device provided by an embodiment of the present disclosure includes: a signal acquisition unit 201, a first noise reduction processing unit 202, and a second noise reduction processing unit 203.
  • the signal acquisition unit 201 is configured to acquire the sound input signal received by the target device.
  • the first noise reduction processing unit 202 is configured to use the external noise signal of the target device to perform noise reduction processing on the sound input signal to obtain the external target sound signal.
  • the second noise reduction processing unit 203 is configured to use the internal noise signal of the target device to perform noise reduction processing on the external target sound signal to obtain the target sound signal.
  • the signal acquisition unit 201 may include:
  • the acquiring subunit is used to acquire the initial sound signal of the external input target device of the target device;
  • the noise reduction subunit is used to perform noise reduction processing on the initial sound signal to obtain the sound input signal.
  • the signal acquisition unit 201 may include:
  • the acquiring subunit is used to acquire the initial sound signal of the external input target device of the target device;
  • the conversion subunit uses the initial sound signal as the sound input signal.
  • the sound input signal received by the target device is first acquired, and then the external noise signal of the target device obtained in advance is used to perform noise reduction processing on the sound input signal to obtain the external target sound signal. Then, the internal noise signal of the target device is used to perform noise reduction processing on the external target sound signal to obtain the target sound signal.
  • the noise reduction device provided by at least one embodiment of the present disclosure can not only avoid the influence of the external noise of the device on the accuracy of speech recognition, but also avoid the influence of the internal noise of the device on the accuracy of the speech recognition of the device, and improve the effect of noise reduction processing and the performance of the device. Intelligence.
  • each unit in the noise reduction device may be hardware, software, firmware, and any feasible combination thereof.
  • the above-mentioned units may be dedicated or general-purpose circuits, chips or devices, etc., or may be a combination of a processor and a memory.
  • the embodiments of the present disclosure do not limit this.
  • each unit in the noise reduction device corresponds to each step of the aforementioned noise reduction method.
  • the specific functions of the noise reduction device please refer to the related description of the noise reduction method. Repeat it again.
  • the components and structure of the noise reduction device shown in FIG. 2 are only exemplary, and not restrictive.
  • the noise reduction device may also include other components and structures as required.
  • FIG. 3 is a schematic block diagram of another noise reduction device provided by at least one embodiment of the present disclosure.
  • the noise reduction device includes a processor 410 and a memory 420.
  • the memory 420 is used to store non-transitory computer readable instructions (for example, one or more computer program modules).
  • the processor 410 is configured to run non-transitory computer-readable instructions, and when the non-transitory computer-readable instructions are executed by the processor 410, one or more steps in the noise reduction method described above can be executed.
  • the memory 420 and the processor 410 may be interconnected by a bus system and/or other forms of connection mechanisms (not shown).
  • the processor 410 may be a central processing unit (CPU), a graphics processing unit (GPU), a digital signal processor (DSP), or other forms of processing units with data processing capabilities and/or program execution capabilities, such as field programmable Gate array (FPGA), etc.; for example, the central processing unit (CPU) can be an X86 or ARM architecture.
  • the processor 410 may be a general-purpose processor or a dedicated processor, and may control other components in the noise reduction device to perform desired functions.
  • the memory 420 may include any combination of one or more computer program products, and the computer program product may include various forms of computer-readable storage media, such as volatile memory and/or nonvolatile memory.
  • Volatile memory may include random access memory (RAM) and/or cache memory (cache), for example.
  • the non-volatile memory may include, for example, read only memory (ROM), hard disk, erasable programmable read only memory (EPROM), portable compact disk read only memory (CD-ROM), USB memory, flash memory, etc.
  • One or more computer program modules may be stored on the computer-readable storage medium, and the processor 410 may run one or more computer program modules to implement various functions of the noise reduction device.
  • the computer-readable storage medium may also store various application programs and various data, various data used and/or generated by the application programs, and the like.
  • At least one embodiment of the present disclosure also provides a device that can reduce noise.
  • FIG. 4 is a schematic block diagram of a device capable of noise reduction provided by at least one embodiment of the present disclosure.
  • An embodiment of the present disclosure provides a device that can achieve noise reduction, including: a first sound acquisition module 301, a second sound acquisition module 302, a housing 303, and a noise processing module 304;
  • the first sound acquisition module 301 is arranged inside the housing 303, and the first sound acquisition module 301 is soundproofed from the outside of the device that can reduce noise;
  • the first sound acquisition module 301 is configured to acquire the internal noise signal inside the device that can achieve noise reduction, and send the internal noise signal to the noise processing module 304;
  • the second sound acquisition module 302 is used to acquire the sound input signal external to the device that can achieve noise reduction, and send the sound input signal to the noise processing module 304; the second sound acquisition module 302 is also used to acquire the noise reduction External noise signal outside the device, and send the external noise signal to the noise processing module 304;
  • the noise processing module 304 is configured to use the external noise signal of the device that can achieve noise reduction to perform noise reduction processing on the sound input signal to obtain an external target sound signal; then use the internal noise signal of the device that can achieve noise reduction to The external target sound signal undergoes noise reduction processing to obtain the target sound signal.
  • the noise processing module 304 uses a fixed beam form or an adaptive filter to perform noise reduction processing.
  • the noise processing module 304 is configured to execute any one of the noise reduction methods provided in the foregoing embodiments. For details, refer to related descriptions.
  • the noise processing module 304 may specifically include a memory and a processor, where the memory is used to store program code and send the program code to the processor, and the processor is used to execute the program code to implement the above-mentioned embodiments. Any of the noise reduction methods.
  • the first sound acquisition module 301 can be set to be highly isolated from the external sound of the device that can achieve noise reduction, such as setting the housing 303 It is a soundproof housing, or a soundproof mechanism for the first sound acquisition module 301 is added to the inside of the housing 303, so that the first sound acquisition module 301 only receives sound signals (ie, internal noise signals) inside the device.
  • sound signals ie, internal noise signals
  • the second sound acquisition module 302 may be used to receive the initial sound signal input from the outside of the device, and perform noise reduction processing on the initial sound signal to obtain the sound input signal.
  • the specific noise reduction method can be selected according to actual needs and is not limited here.
  • a first sound acquisition module 301 is provided inside the housing 303 of the device.
  • the first sound acquisition module 301 is soundproofed from the outside of the device, and the first sound acquisition module 301 can be used to acquire the internal Noise signal, and use the second sound acquisition module 302 to acquire the external noise signal and sound input signal external to the device, and send the acquired sound input signal, external noise signal, and internal noise signal to the noise processing module 303, so that the noise processing module 303 According to the external noise signal and the internal noise signal, the noise input signal is processed to obtain the target sound signal, which avoids the influence of the internal noise of the device on the accuracy of the device's voice recognition, and improves the effect of noise reduction processing and the intelligence of the device.
  • the device can be the aforementioned target device, and can be any device that can perform voice recording, such as a recording device, a device with voice recognition, and a device that can be voice controlled, including but not limited to voice recorders, camcorders, etc. , Smart phones, smart speakers, smart home devices (such as sweeping robots, air purifiers, air conditioners, fans, etc.), smart translators, etc., are not limited here.
  • the device may also only include the first sound acquisition module 301, the second sound acquisition module 302, and the housing 303, but not the noise processing module 304.
  • the noise processing module 304 is, for example, provided in other devices connected to the device signal.
  • the device provides external noise signals, internal noise signals, and sound input signals to the noise processing module 304 in other devices, thereby achieving noise reduction and obtaining target sounds. signal.
  • the setting modes and functions of the various modules of the device can be adjusted accordingly, which is not limited by the embodiments of the present disclosure.
  • the device may also include more modules and components to achieve more comprehensive functions, which may be determined according to actual requirements, which are not limited in the embodiments of the present disclosure.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computational Linguistics (AREA)
  • Health & Medical Sciences (AREA)
  • Audiology, Speech & Language Pathology (AREA)
  • Human Computer Interaction (AREA)
  • Physics & Mathematics (AREA)
  • Acoustics & Sound (AREA)
  • Multimedia (AREA)
  • Soundproofing, Sound Blocking, And Sound Damping (AREA)
  • Circuit For Audible Band Transducer (AREA)

Abstract

一种降噪方法、降噪装置及可实现降噪的设备，该方法包括：获取目标设备接收到的声音输入信号（S101）；利用目标设备的外部噪声信号，对声音输入信号进行降噪处理，得到外部目标声音信号（S102）；利用目标设备的内部噪声信号，对外部目标声音信号进行降噪处理，得到目标声音信号（S103）。该方法不仅能够避免设备外部噪声对语音识别准确性的影响，还避免了设备内部噪声对设备语音识别准确性的影响，提高了降噪处理的效果和设备的智能性。

Description

降噪方法、降噪装置及可实现降噪的设备
本申请要求于2019年5月29日递交的中国专利申请第201910457370.0号的优先权，该中国专利申请的全文以引入的方式并入以作为本申请的一部分。
技术领域
本公开的实施例涉及一种降噪方法、降噪装置及可实现降噪的设备。
背景技术
随着人工智能技术的发展，越来越多的智能语音设备出现在人们的生活中。为了保证语音识别的准确性并提高设备的智能性，智能设备需要对接收到的语音进行降噪处理。
发明内容
本公开至少一个实施例提供一种降噪方法，包括：获取目标设备接收到的声音输入信号；利用所述目标设备的外部噪声信号，对所述声音输入信号进行降噪处理，得到外部目标声音信号；利用所述目标设备的内部噪声信号，对所述外部目标声音信号进行降噪处理，得到所述目标声音信号。
例如，在本公开一实施例提供的方法中，所述获取目标设备接收到的声音输入信号，包括：获取所述目标设备的外部输入所述目标设备的初始声音信号；对所述初始声音信号进行降噪处理，得到所述声音输入信号。
例如，在本公开一实施例提供的方法中，所述获取目标设备接收到的声音输入信号，包括：获取所述目标设备的外部输入所述目标设备的初始声音信号；将所述初始声音信号作为所述声音输入信号。
例如，在本公开一实施例提供的方法中，所述降噪处理利用固定波束形式或者自适应滤波器进行。
例如，在本公开一实施例提供的方法中，所述内部噪声信号经由所述目标设备内部设置的第一声音获取模块获得；所述第一声音获取模块与所述目标设备的外部隔音。
例如，在本公开一实施例提供的方法中，所述目标设备的外部套设有壳体，所述壳体隔音。
本公开至少一个实施例还提供一种降噪方法，包括：获取目标设备接收到的声音输入信号；利用所述目标设备的内部噪声信号，对所述声音输入信号进行降噪处理，得到第一声音信号；利用所述目标设备的外部噪声信号，对所述第一声音信号进行降噪处理，得到目标声音信号。
本公开至少一个实施例提供一种降噪方法，包括：获取目标设备接收到的声音输入信号；利用所述目标设备的外部噪声信号和内部噪声信号，在同一降噪操作中对所述声音输入信号进行降噪处理，得到目标声音信号。
本公开至少一个实施例提供一种降噪方法，包括：获取目标设备接收到的声音输入信号；将所述目标设备的外部噪声信号作为唯一的参考信号，对所述声音输入信号进行降噪处理，得到目标声音信号。
本公开至少一个实施例提供一种降噪方法，包括：获取目标设备接收到的声音输入信号；将所述目标设备的内部噪声信号作为唯一的参考信号，对所述声音输入信号进行降噪处理，得到目标声音信号。
本公开至少一个实施例提供一种降噪装置，包括：信号获取单元、第一降噪处理单元和第二降噪处理单元；其中，所述信号获取单元配置为获取目标设备接收到的声音输入信号；所述第一降噪处理单元配置为，利用所述目标设备的外部噪声信号，对所述声音输入信号进行降噪处理，得到外部目标声音信号；所述第二降噪处理单元配置为，利用所述目标设备的内部噪声信号，对所述外部目标声音信号进行降噪处理，得到所述目标声音信号。
本公开至少一个实施例提供一种降噪装置，包括：处理器；存储器，包括一个或多个计算机程序模块；其中，所述一个或多个计算机程序模块被存储在所述存储器中并被配置为由所述处理器执行，所述一个或多个计算机程序模块包括用于实现本公开任一实施例所述的降噪方法的指令。
本公开至少一个实施例提供一种可实现降噪的设备，包括：第一声音获取模块、第二声音获取模块、壳体和噪声处理模块；其中，所述第一声音获取模块设置在所述壳体的内部，所述第一声音获取模块与所述设备的外部隔音；所述第一声音获取模块配置为获取所述设备内部的内部噪声信号，并将所述内部噪声信号发送至所述噪声处理模块；所述第二声音获取模块配置为获取所述设备外部的声音输入信号，并将所述声音输入信号发送至所述噪声 处理模块；所述第二声音获取模块还配置为获取所述设备外部的外部噪声信号，并将所述外部噪声信号发送至所述噪声处理模块；所述噪声处理模块配置为利用所述外部噪声信号，对所述声音输入信号进行降噪处理，得到外部目标声音信号，再利用所述内部噪声信号，对所述外部目标声音信号进行降噪处理，得到目标声音信号。
例如，在本公开一实施例提供的设备中，所述壳体隔音。
例如，在本公开一实施例提供的设备中，所述第二声音获取模块还配置为接收所述设备外部输入的初始声音信号，并对所述初始声音信号进行降噪处理，得到所述声音输入信号。
例如，在本公开一实施例提供的设备中，所述噪声处理模块利用固定波束形式或者自适应滤波器进行降噪处理。
附图说明
为了更清楚地说明本公开实施例或现有技术中的技术方案，下面将对实施例或现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公开中记载的一些实施例，对于本领域普通技术人员来讲，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它的附图。
图1为本公开至少一个实施例提供的一种降噪方法的流程示意图；
图2为本公开至少一个实施例提供的一种降噪装置的示意框图；
图3为本公开至少一个实施例提供的另一种降噪装置的示意框图；以及
图4为本公开至少一个实施例提供的一种可实现降噪的设备的示意框图。
具体实施方式
为了使本技术领域的人员更好地理解本公开的方案，下面将结合本公开实施例中的附图，对本公开实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅是本公开的一部分实施例，而不是全部的实施例。基于本公开中的实施例，本领域普通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，都属于本公开保护的范围。
应当理解，在本公开中，“至少一个(项)”是指一个或者多个，“多个”是指两个或两个以上。“和/或”，用于描述关联对象的关联关系，表示可以存 在三种关系，例如，“A和/或B”可以表示：只存在A，只存在B以及同时存在A和B三种情况，其中A，B可以是单数或者复数。字符“/”一般表示前后关联对象是一种“或”的关系。“以下至少一项(个)”或其类似表达，是指这些项中的任意组合，包括单项(个)或复数项(个)的任意组合。例如，a，b或c中的至少一项(个)，可以表示：a，b，c，“a和b”，“a和c”，“b和c”，或“a和b和c”，其中a，b，c可以是单个，也可以是多个。
除非另外定义，本公开使用的技术术语或者科学术语应当为本公开所属领域内具有一般技能的人士所理解的通常意义。本公开中使用的“第一”、“第二”以及类似的词语并不表示任何顺序、数量或者重要性，而只是用来区分不同的组成部分。同样，“一个”、“一”或者“该”等类似词语也不表示数量限制，而是表示存在至少一个。“包括”或者“包含”等类似的词语意指出现该词前面的元件或者物件涵盖出现在该词后面列举的元件或者物件及其等同，而不排除其他元件或者物件。“连接”或者“相连”等类似的词语并非限定于物理的或者机械的连接，而是可以包括电性的连接，不管是直接的还是间接的。“上”、“下”、“左”、“右”等仅用于表示相对位置关系，当被描述对象的绝对位置改变后，则该相对位置关系也可能相应地改变。
智能设备在实现语音识别时，一般会对输入的语音进行降噪处理，以保证语音识别的准确性。然而，本公开的发明人在研究中发现，当智能设备内部噪声较大时，设备内部噪声对语音识别的干扰较大，现有的智能设备仍然存在降噪效果不佳的问题，无法保证语音识别的准确性，导致设备的智能化程度低。
因此，本公开的实施例提供了一种降噪方法、降噪装置及可实现降噪的设备，不仅能够消除设备外部的噪声，还可以利用设备内部设置的第二声音获取模块获取设备内部的内部噪声信号，以消除设备内部的噪声，提高噪声消除的效果和语音识别的准确性，保证设备的高智能化。
例如，在本公开至少一个实施例中，首先获取目标设备接收到的声音输入信号，再利用预先获得的目标设备外部噪声信号，对声音输入信号进行降噪处理，得到外部目标声音信号。然后，利用目标设备的内部噪声信号，对外部目标声音信号进行降噪处理，得到目标声音信号。本公开至少一个实施例提供的降噪方法不仅能够避免设备外部噪声对语音识别准确性的影响，还避免了设备内部噪声对设备语音识别准确性的影响，提高了降噪处理的效果 和设备的智能性。
基于上述思想，为使本公开的上述目的、特征和优点能够更加明显易懂，下面结合附图对本公开的具体实施方式做详细的说明。
参见图1，该图为本公开至少一个实施例提供的一种降噪方法的流程示意图。
本公开至少一个实施例提供的一种降噪方法，包括：
步骤S101：获取目标设备接收到的声音输入信号。
需要说明的是，目标设备具体可以是任意一种可进行语音录入的设备，如录音设备、具备语音识别的设备和可语音控制的设备等，包括但不限于录音笔、摄录机、智能手机、智能音响、智能家居设备(如扫地机器人、空气净化器、空调、风扇等)、智能翻译机等，在此不进行限定。
在本公开的实施例中，接收到的声音输入信号不仅包括外部噪声信号，还包括期望识别的目标声音信号(如用户录入的语音指令等)，本公开的实施例对目标声音信号的形式和声音输入信号的获得方式不进行限定。
作为一个示例，声音输入信号可以利用设置在目标设备上或者目标设备外部的第二声音获取模块获得。在具体实施时，第二声音获取模块可以包括一个或多个收音设备(如任意结构的麦克风等)。在实际应用中，可以根据实际情况及需要，对第二声音获取模块所包括的收音设备在目标设备上的位置及排列方式进行具体设定。例如，第二声音获取模块可以是外接的麦克风，也可以是设置在目标设备壳体上的麦克风，第二声音获取模块还可以包括多个以环绕方式排列在目标设备壳体上的麦克风，第二声音获取模块还可以为其他任意适用的设置方式，这里不再一一列举。
在本公开实施例一些可能的实现方式中，为了提高降噪的效果，可以对实际接收到的声音信号(即初始声音信号)进行预降噪。则，步骤S101可以包括如下操作：
步骤S1011：获取目标设备的外部输入目标设备的初始声音信号；
步骤S1012：对初始声音信号进行降噪处理，得到声音输入信号。
可以理解的是，初始声音信号可以为目标设备直接接收到的原始声音信号。在实际应用中，可以采用任意一种降噪处理算法对初始声音信号进行降噪，例如该降噪处理可以利用固定波束形式进行。所谓固定波束形成是指波束形成过程中，滤波器权值固定不变。一旦麦克风阵列的几何形状、目标方 向确定，其波束模式特性也将确定。该方法具有简单、运算量低的优点，这里不再赘述。
需要说明的是，本公开的实施例不限于此，在另一些示例中，步骤S101可以包括如下操作：
步骤S1013：获取目标设备的外部输入目标设备的初始声音信号；
步骤S1014：将初始声音信号作为声音输入信号。
在该示例中，初始声音信号可以为目标设备直接接收到的原始声音信号，直接将该初始声音信号作为上述声音输入信号，而省略对该初始声音信号进行降噪的操作。通过这种方式，可以在降噪要求较低时，有效提高处理效率，减少运算量。
步骤S102：利用目标设备的外部噪声信号，对声音输入信号进行降噪处理，得到外部目标声音信号。
在本公开的实施例中，外部噪声信号可以是目标设备外部的干扰信号，如环境噪声等。将外部噪声信号作为消除目标设备外部噪声的参考信号进行降噪处理，可以消除目标设备外部噪声对语音录入或识别准确度的影响。可以理解的是，外部噪声信号也可以利用第二声音获取模块获得。例如，当用户未向目标设备录入语音时，第二声音获取模块接收到的声音信号可以作为外部噪声信号。
在实际应用中，可以利用任意一种降噪算法进行降噪处理，例如固定波束或自适应等，这里不进行限定。作为一个示例，可以利用最小均方(least mean square，LMS)自适应滤波器、归一化最小均方(normalized least mean square，NLMS)自适应滤波器、递归式最小均方(recursive least square，RLS)自适应滤波器等进行降噪处理。
例如，LMS自适应滤波器的原理是：利用前一刻已获得的滤波器参数，自动调节当前滤波器参数，使得输入信号序列x(n)和期望输出信号d(n)逼近，以适应信号和噪声未知的或随机变化的统计特性，从而实现最优滤波。
输入信号序列为x(n)和期望输出信号为d(n)的误差信号e(n)如下式(1)，
Figure PCTCN2019129930-appb-000001
式中，ω i为权系数，n∈[1，M]。
LMS算法的本质就是寻找最优的权系数，经过一系列推导，最终可得到下式(2)，
ω k(n)＝ω k(n-1)+μe(n)x(n)          (2)
式中，μ为收敛因子，μ的大小决定着系统的稳定性和收敛速度。其他自适应滤波器的原理与LMS类似，区别仅在于使得输入信号序列x(n)和期望输出信号d(n)逼近的标准不同，这里不再一一赘述。
步骤S103：利用目标设备的内部噪声信号，对外部目标声音信号进行降噪处理，得到目标声音信号。
在本公开的实施例中，内部噪声信号可以是目标设备内部的干扰信号，如目标设备的运行噪声等。将内部噪声信号作为消除目标设备内部噪声的参考信号进行降噪，可以消除目标设备内部噪声(如电机运行噪声等)对语音录入或识别准确性的影响。由于首先利用外部噪声信号作为降噪的参考信号进行降噪，降低了外部噪声的影响，再利用内部噪声信号作为降噪的参考信号进行二次降噪，又降低了内部噪声的影响，从而提高了降噪的效果和准确度，保证了目标设备的智能性。
在实际应用中，内部噪声信号可以利用任意一种收音设备获得。在本公开实施例一些可能的实现方式中，内部噪声信号可以经由目标设备内部设置的第一声音获取模块获得，为了保证降噪的准确性，该第一声音获取模块与目标设备的外部隔音。作为一个示例，可以将目标设备外套设的壳体设置为隔音壳体，或者，在目标设备的壳体内部增设对第一声音获取模块的隔音机构等，以使第一声音获取模块仅接收到目标设备内部的声音信号(即内部噪声信号)。
在具体实施时，第一声音获取模块可以包括至少一个收音设备(如任意结构的麦克风等)，该至少一个收音设备在目标设备壳体内部的位置及排列方式可以根据实际情况具体设定，例如第一声音获取模块可以包括多个环绕排列设置在目标设备内部的麦克风，这里不进行限定，也不再一一列举。
可以理解的是，与步骤S102类似，在实际应用中，可以利用任意一种降噪算法对外部目标声音信号进行降噪处理，例如固定波束或自适应等，这里不进行限定。
在本公开至少一个实施例提供的降噪方法中，首先获取目标设备接收到的声音输入信号，再利用预先获得的目标设备的外部噪声信号，对声音输入信号进行降噪处理，得到外部目标声音信号。然后，利用目标设备的内部噪声信号，对外部目标声音信号进行降噪处理，得到目标声音信号。本公开至 少一个实施例提供的降噪方法不仅能够避免设备外部噪声对语音识别准确性的影响，还避免了设备内部噪声对设备语音识别准确性的影响，提高了降噪处理的效果和设备的智能性。
本公开至少一个实施例还提供一种降噪方法，该降噪方法包括如下操作。
步骤S11：获取目标设备接收到的声音输入信号；
步骤S12：利用目标设备的内部噪声信号，对声音输入信号进行降噪处理，得到第一声音信号；
步骤S13：利用目标设备的外部噪声信号，对第一声音信号进行降噪处理，得到目标声音信号。
在该实施例中，步骤S12中的降噪处理是以内部噪声信号作为参考信号，步骤S13中的降噪处理是以外部噪声信号作为参考信号，该实施例的处理顺序与图1所示的降噪方法中的处理顺序不同。通过这种方式，可以在内部噪声较大时首先将内部噪声信号作为参考信号进行降噪处理，从而提高最终语音识别结果的准确性。步骤S12和步骤S13中进行降噪处理的具体方式可以参考前文内容，此处不再赘述。
本公开至少一个实施例还提供一种降噪方法，该降噪方法包括如下操作。
步骤S21：获取目标设备接收到的声音输入信号；
步骤S22：利用目标设备的外部噪声信号和内部噪声信号，在同一降噪操作中对声音输入信号进行降噪处理，得到目标声音信号。
在该实施例中，步骤S22中的降噪处理是以外部噪声信号和内部噪声信号同时作为参考信号，也即是，这两个参考信号在同一降噪操作中使用，以实现降噪处理，这与图1所示的降噪方法中的依序分步处理方式不同。通过这种方式，可以简化处理算法的逻辑。关于降噪处理的具体方式可以参考前文内容，此处不再赘述。
本公开至少一个实施例还提供一种降噪方法，该降噪方法包括如下操作。
步骤S31：获取目标设备接收到的声音输入信号；
步骤S32：将目标设备的外部噪声信号作为唯一的参考信号，对声音输入信号进行降噪处理，得到目标声音信号。
在该实施例中，仅仅使用外部噪声信号作为唯一的参考信号，而不再使用内部噪声信号，可以在内部噪声较小时简化操作，提高处理效率，减少运算量。同时，由于无需在相应的设备中设置用于获取内部噪声的声音获取模 块，因此可以简化设备结构。关于降噪处理的具体方式可以参考前文内容，此处不再赘述。
本公开至少一个实施例还提供一种降噪方法，该降噪方法包括如下操作。
步骤S41：获取目标设备接收到的声音输入信号；
步骤S42：将目标设备的内部噪声信号作为唯一的参考信号，对声音输入信号进行降噪处理，得到目标声音信号。
在该实施例中，仅仅使用内部噪声信号作为唯一的参考信号，而不再使用外部噪声信号，可以在内部噪声较大、外部噪声较小时简化操作，提高处理效率，减少运算量。关于降噪处理的具体方式可以参考前文内容，此处不再赘述。
本公开至少一个实施例还提供了一种降噪装置。
参见图2，该图为本公开至少一个实施例提供的一种降噪装置的示意框图。
本公开实施例提供的一种降噪装置，包括：信号获取单元201、第一降噪处理单元202和第二降噪处理单元203。
信号获取单元201，用于获取目标设备接收到的声音输入信号。
第一降噪处理单元202，用于利用目标设备的外部噪声信号，对声音输入信号进行降噪处理，得到外部目标声音信号。
第二降噪处理单元203，用于利用目标设备的内部噪声信号，对外部目标声音信号进行降噪处理，得到目标声音信号。
在本公开实施例一些可能的实现方式中，信号获取单201，可以包括：
获取子单元，用于获取目标设备的外部输入目标设备的初始声音信号；
降噪子单元，用于对初始声音信号进行降噪处理，得到声音输入信号。
例如，在本公开实施例另一些可能的实现方式中，信号获取单201，可以包括：
获取子单元，用于获取目标设备的外部输入目标设备的初始声音信号；
转换子单元，将初始声音信号作为声音输入信号。
在本公开至少一个实施例提供的降噪装置中，首先获取目标设备接收到的声音输入信号，再利用预先获得的目标设备的外部噪声信号，对声音输入信号进行降噪处理，得到外部目标声音信号。然后，利用目标设备的内部噪声信号，对外部目标声音信号进行降噪处理，得到目标声音信号。本公开至 少一个实施例提供的降噪装置不仅能够避免设备外部噪声对语音识别准确性的影响，还避免了设备内部噪声对设备语音识别准确性的影响，提高了降噪处理的效果和设备的智能性。
需要说明的是，本公开的实施例中，降噪装置中的各个单元可以为硬件、软件、固件以及它们的任意可行的组合。例如，上述各个单元可以为专用或通用的电路、芯片或装置等，也可以为处理器和存储器的结合。关于上述各个单元的具体实现形式，本公开的实施例对此不作限制。
需要说明的是，本公开的实施例中，降噪装置中的各个单元与前述的降噪方法的各个步骤对应，关于降噪装置的具体功能可以参考关于降噪方法的相关描述，此处不再赘述。图2所示的降噪装置的组件和结构只是示例性的，而非限制性的，根据需要，该降噪装置还可以包括其他组件和结构。
图3为本公开至少一个实施例提供的另一种降噪装置的示意框图。如图3所示，该降噪装置包括处理器410和存储器420。存储器420用于存储非暂时性计算机可读指令(例如一个或多个计算机程序模块)。处理器410用于运行非暂时性计算机可读指令，非暂时性计算机可读指令被处理器410运行时可以执行上文所述的降噪方法中的一个或多个步骤。存储器420和处理器410可以通过总线系统和/或其它形式的连接机构(未示出)互连。
例如，处理器410可以是中央处理单元(CPU)、图形处理单元(GPU)、数字信号处理器(DSP)或者具有数据处理能力和/或程序执行能力的其它形式的处理单元，例如现场可编程门阵列(FPGA)等；例如，中央处理单元(CPU)可以为X86或ARM架构等。处理器410可以为通用处理器或专用处理器，可以控制降噪装置中的其它组件以执行期望的功能。
例如，存储器420可以包括一个或多个计算机程序产品的任意组合，计算机程序产品可以包括各种形式的计算机可读存储介质，例如易失性存储器和/或非易失性存储器。易失性存储器例如可以包括随机存取存储器(RAM)和/或高速缓冲存储器(cache)等。非易失性存储器例如可以包括只读存储器(ROM)、硬盘、可擦除可编程只读存储器(EPROM)、便携式紧致盘只读存储器(CD-ROM)、USB存储器、闪存等。在计算机可读存储介质上可以存储一个或多个计算机程序模块，处理器410可以运行一个或多个计算机程序模块，以实现降噪装置的各种功能。在计算机可读存储介质中还可以存储各种应用程序和各种数据以及应用程序使用和/或产生的各种数据等。
需要说明的是，本公开的实施例中，降噪装置的具体功能和技术效果可以参考上文中关于降噪方法的描述，此处不再赘述。
本公开至少一个实施例还提供了一种可实现降噪的设备。
参见图4，该图为本公开至少一个实施例提供的一种可实现降噪的设备的示意框图。
本公开实施例提供的一种可实现降噪的设备，包括：第一声音获取模块301、第二声音获取模块302、壳体303和噪声处理模块304；
第一声音获取模块301设置在壳体303的内部，第一声音获取模块301与该可实现降噪的设备的外部隔音；
第一声音获取模块301，用于获取该可实现降噪的设备内部的内部噪声信号，并将内部噪声信号发送至噪声处理模块304；
第二声音获取模块302，用于获取该可实现降噪的设备外部的声音输入信号，并将声音输入信号发送至噪声处理模块304；第二声音获取模块302还用于获取该可实现降噪的设备外部的外部噪声信号，并将外部噪声信号发送至噪声处理模块304；
噪声处理模块304，用于利用该可实现降噪的设备的外部噪声信号，对声音输入信号进行降噪处理，得到外部目标声音信号；再利用该可实现降噪的设备的内部噪声信号，对外部目标声音信号进行降噪处理，得到目标声音信号。
作为一个示例，噪声处理模块304利用固定波束形式或者自适应滤波器进行降噪处理。
可以理解的是，第一声音获取模块301和第二声音获取模块302与上述方法实施例中所述的第一声音获取模块和第二声音获取模块类似，具体说明可参照上述方法实施例中的相关内容即可，这里不再赘述。噪声处理模块304用于执行上述实施例提供的降噪方法中的任意一种，具体可以参照相关说明。
在实际应用中，噪声处理模块304具体可以包括存储器和处理器，其中存储器用于存储程序代码，并将该程序代码发送至处理器，该处理器用于执行该程序代码以实现如上述实施例提供的降噪方法中的任意一种。
在本公开实施例一些可能的实现方式中，为了进一步提高降噪的效果，可以将第一声音获取模块301设置为与该可实现降噪的设备的外部声音高隔离，如将壳体303设置为隔音壳体，或者，在壳体303的内部增设对第一声 音获取模块301的隔音机构等，以使第一声音获取模块301仅接收到设备内部的声音信号(即内部噪声信号)。
在本公开实施例一些可能的实现方式中，为了提高降噪的效果，第二声音获取模块302可用于接收设备的外部输入的初始声音信号，并对初始声音信号进行降噪处理，得到声音输入信号。具体降噪方法可以根据实际需要选取，这里不进行限定。
在本公开的实施例中，在设备的壳体303内部设置有第一声音获取模块301，第一声音获取模块301与设备的外部隔音，则可利用第一声音获取模块301获取设备内部的内部噪声信号，并利用第二声音获取模块302获取设备外部的外部噪声信号和声音输入信号，将获取到的声音输入信号、外部噪声信号和内部噪声信号发送至噪声处理模块303，使得噪声处理模块303根据外部噪声信号和内部噪声信号，对声音输入信号进行降噪处理得到目标声音信号，避免了设备内部噪声对设备语音识别准确性的影响，提高了降噪处理的效果和设备的智能性。
例如，该设备可以为前述的目标设备，可以是任意一种可进行语音录入的设备，如录音设备、具备语音识别的设备和可语音控制的设备等，包括但不限于录音笔、摄录机、智能手机、智能音响、智能家居设备(如扫地机器人、空气净化器、空调、风扇等)、智能翻译机等，在此不进行限定。
例如，在一些示例中，该设备也可以仅包括第一声音获取模块301、第二声音获取模块302和壳体303，而不包括噪声处理模块304。噪声处理模块304例如设置在与该设备信号连接的其他装置中，该设备将外部噪声信号、内部噪声信号和声音输入信号提供给其他装置中的噪声处理模块304，从而实现降噪并获得目标声音信号。
需要说明的是，本公开的实施例中，根据上文所述的降噪方法的不同操作步骤，该设备的各个模块的设置方式和功能可以相应调整，本公开的实施例对此不作限制。该设备还可以包括更多的模块和部件，以实现更加全面的功能，这可以根据实际需求而定，本公开的实施例对此不作限制。
需要说明的是，本说明书中各个实施例采用递进的方式描述，每个实施例重点说明的都是与其他实施例的不同之处，各个实施例之间相同相似部分互相参见即可。对于实施例公开的系统或装置而言，由于其与实施例公开的方法相对应，所以描述比较简单，相关之处参见方法部分说明即可。
有以下几点需要说明：
(1)本公开实施例附图只涉及到本公开实施例涉及到的结构，其他结构可参考通常设计。
(2)在不冲突的情况下，本公开的实施例及实施例中的特征可以相互组合以得到新的实施例。
以上所述，仅为本公开的具体实施方式，但本公开的保护范围并不局限于此，本公开的保护范围应以所述权利要求的保护范围为准。

Claims (16)

  1. 一种降噪方法，包括：
    获取目标设备接收到的声音输入信号；
    利用所述目标设备的外部噪声信号，对所述声音输入信号进行降噪处理，得到外部目标声音信号；
    利用所述目标设备的内部噪声信号，对所述外部目标声音信号进行降噪处理，得到所述目标声音信号。
  2. 根据权利要求1所述的方法，其中，所述获取目标设备接收到的声音输入信号，包括：
    获取所述目标设备的外部输入所述目标设备的初始声音信号；
    对所述初始声音信号进行降噪处理，得到所述声音输入信号。
  3. 根据权利要求1所述的方法，其中，所述获取目标设备接收到的声音输入信号，包括：
    获取所述目标设备的外部输入所述目标设备的初始声音信号；
    将所述初始声音信号作为所述声音输入信号。
  4. 根据权利要求1-3任一项所述的方法，其中，所述降噪处理利用固定波束形式或者自适应滤波器进行。
  5. 根据权利要求1-3任一项所述的方法，其中，所述内部噪声信号经由所述目标设备内部设置的第一声音获取模块获得；
    所述第一声音获取模块与所述目标设备的外部隔音。
  6. 根据权利要求5所述的方法，其中，所述目标设备的外部套设有壳体，所述壳体隔音。
  7. 一种降噪方法，包括：
    获取目标设备接收到的声音输入信号；
    利用所述目标设备的内部噪声信号，对所述声音输入信号进行降噪处理，得到第一声音信号；
    利用所述目标设备的外部噪声信号，对所述第一声音信号进行降噪处理，得到目标声音信号。
  8. 一种降噪方法，包括：
    获取目标设备接收到的声音输入信号；
    利用所述目标设备的外部噪声信号和内部噪声信号，在同一降噪操作中对所述声音输入信号进行降噪处理，得到目标声音信号。
  9. 一种降噪方法，包括：
    获取目标设备接收到的声音输入信号；
    将所述目标设备的外部噪声信号作为唯一的参考信号，对所述声音输入信号进行降噪处理，得到目标声音信号。
  10. 一种降噪方法，包括：
    获取目标设备接收到的声音输入信号；
    将所述目标设备的内部噪声信号作为唯一的参考信号，对所述声音输入信号进行降噪处理，得到目标声音信号。
  11. 一种降噪装置，包括：信号获取单元、第一降噪处理单元和第二降噪处理单元；其中，
    所述信号获取单元配置为获取目标设备接收到的声音输入信号；
    所述第一降噪处理单元配置为，利用所述目标设备的外部噪声信号，对所述声音输入信号进行降噪处理，得到外部目标声音信号；
    所述第二降噪处理单元配置为，利用所述目标设备的内部噪声信号，对所述外部目标声音信号进行降噪处理，得到所述目标声音信号。
  12. 一种降噪装置，包括：
    处理器；
    存储器，包括一个或多个计算机程序模块；
    其中，所述一个或多个计算机程序模块被存储在所述存储器中并被配置为由所述处理器执行，所述一个或多个计算机程序模块包括用于实现权利要求1-10任一项所述的降噪方法的指令。
  13. 一种可实现降噪的设备，包括：第一声音获取模块、第二声音获取模块、壳体和噪声处理模块；其中，
    所述第一声音获取模块设置在所述壳体的内部，所述第一声音获取模块与所述设备的外部隔音；所述第一声音获取模块配置为获取所述设备内部的内部噪声信号，并将所述内部噪声信号发送至所述噪声处理模块；
    所述第二声音获取模块配置为获取所述设备外部的声音输入信号，并将所述声音输入信号发送至所述噪声处理模块；所述第二声音获取模块还配置为获取所述设备外部的外部噪声信号，并将所述外部噪声信号发送至所述噪 声处理模块；
    所述噪声处理模块配置为利用所述外部噪声信号，对所述声音输入信号进行降噪处理，得到外部目标声音信号，再利用所述内部噪声信号，对所述外部目标声音信号进行降噪处理，得到目标声音信号。
  14. 根据权利要求13所述的设备，其中，所述壳体隔音。
  15. 根据权利要求13所述的设备，其中，所述第二声音获取模块还配置为接收所述设备外部输入的初始声音信号，并对所述初始声音信号进行降噪处理，得到所述声音输入信号。
  16. 根据权利要求13-15任一项所述的设备，其中，所述噪声处理模块利用固定波束形式或者自适应滤波器进行降噪处理。
PCT/CN2019/129930 2019-05-29 2019-12-30 降噪方法、降噪装置及可实现降噪的设备 WO2020238203A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201910457370.0A CN110164425A (zh) 2019-05-29 2019-05-29 一种降噪方法、装置及可实现降噪的设备
CN201910457370.0 2019-05-29

Publications (1)

Publication Number Publication Date
WO2020238203A1 true WO2020238203A1 (zh) 2020-12-03

Family

ID=67629691

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/129930 WO2020238203A1 (zh) 2019-05-29 2019-12-30 降噪方法、降噪装置及可实现降噪的设备

Country Status (2)

Country Link
CN (1) CN110164425A (zh)
WO (1) WO2020238203A1 (zh)

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110164425A (zh) * 2019-05-29 2019-08-23 北京声智科技有限公司 一种降噪方法、装置及可实现降噪的设备
CN111341304A (zh) * 2020-02-28 2020-06-26 广州国音智能科技有限公司 一种基于gan的说话人语音特征训练方法、装置和设备
CN114155871A (zh) * 2021-12-16 2022-03-08 科大讯飞股份有限公司 一种语音降噪方法、装置、存储介质及设备
CN117935813A (zh) * 2024-03-21 2024-04-26 武汉理工大学 一种声纹识别方法及系统

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP1306832A1 (en) * 2000-06-09 2003-05-02 Japan Science and Technology Corporation Robot acoustic device and robot acoustic system
CN104952458A (zh) * 2015-06-09 2015-09-30 广州广电运通金融电子股份有限公司 一种噪声抑制方法、装置及系统
CN106388700A (zh) * 2016-06-06 2017-02-15 北京小米移动软件有限公司 用于自动清洁设备的主动降噪装置和自动清洁设备
CN107277690A (zh) * 2017-08-02 2017-10-20 北京地平线信息技术有限公司 声音处理方法、装置和电子设备
CN108682428A (zh) * 2018-08-27 2018-10-19 珠海市微半导体有限公司 机器人语音控制系统和机器人对语音信号的处理方法
CN108847221A (zh) * 2018-06-19 2018-11-20 Oppo广东移动通信有限公司 语音识别方法、装置、存储介质及电子设备
CN109737521A (zh) * 2018-03-07 2019-05-10 北京三五二环保科技有限公司 具有语音控制功能的空气净化器
CN110164425A (zh) * 2019-05-29 2019-08-23 北京声智科技有限公司 一种降噪方法、装置及可实现降噪的设备

Family Cites Families (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP5713771B2 (ja) * 2011-04-11 2015-05-07 株式会社オーディオテクニカ 防音送話器
JP6069830B2 (ja) * 2011-12-08 2017-02-01 ソニー株式会社 耳孔装着型収音装置、信号処理装置、収音方法
JP6069829B2 (ja) * 2011-12-08 2017-02-01 ソニー株式会社 耳孔装着型収音装置、信号処理装置、収音方法
CN104424953B (zh) * 2013-09-11 2019-11-01 华为技术有限公司 语音信号处理方法与装置
US9401158B1 (en) * 2015-09-14 2016-07-26 Knowles Electronics, Llc Microphone signal fusion
US9922637B2 (en) * 2016-07-11 2018-03-20 Microsoft Technology Licensing, Llc Microphone noise suppression for computing device
CN206713019U (zh) * 2017-02-10 2017-12-05 万魔声学科技有限公司 主动降噪装置和耳机
CN107452368B (zh) * 2017-08-14 2021-05-07 海尔优家智能科技（北京）有限公司 一种家电设备的降噪方法及装置
CN108088064B (zh) * 2018-01-15 2023-05-23 青岛理工大学 设于通风管口且能够声学交互的有源降噪装置及控制方法
CN208905198U (zh) * 2018-09-30 2019-05-24 科大讯飞股份有限公司 具有全隔离降噪电路的电子设备
CN109243453A (zh) * 2018-10-29 2019-01-18 珠海格力电器股份有限公司 语音信息的处理方法及装置、存储介质、电子装置
CN109599107A (zh) * 2018-12-07 2019-04-09 珠海格力电器股份有限公司 一种语音识别的方法、装置及计算机存储介质
CN210667779U (zh) * 2019-05-29 2020-06-02 北京声智科技有限公司 一种可实现降噪的设备

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP1306832A1 (en) * 2000-06-09 2003-05-02 Japan Science and Technology Corporation Robot acoustic device and robot acoustic system
CN104952458A (zh) * 2015-06-09 2015-09-30 广州广电运通金融电子股份有限公司 一种噪声抑制方法、装置及系统
CN106388700A (zh) * 2016-06-06 2017-02-15 北京小米移动软件有限公司 用于自动清洁设备的主动降噪装置和自动清洁设备
CN107277690A (zh) * 2017-08-02 2017-10-20 北京地平线信息技术有限公司 声音处理方法、装置和电子设备
CN109737521A (zh) * 2018-03-07 2019-05-10 北京三五二环保科技有限公司 具有语音控制功能的空气净化器
CN108847221A (zh) * 2018-06-19 2018-11-20 Oppo广东移动通信有限公司 语音识别方法、装置、存储介质及电子设备
CN108682428A (zh) * 2018-08-27 2018-10-19 珠海市微半导体有限公司 机器人语音控制系统和机器人对语音信号的处理方法
CN110164425A (zh) * 2019-05-29 2019-08-23 北京声智科技有限公司 一种降噪方法、装置及可实现降噪的设备

Also Published As

Publication number Publication date
CN110164425A (zh) 2019-08-23

Similar Documents

Publication Publication Date Title
WO2020238203A1 (zh) 降噪方法、降噪装置及可实现降噪的设备
US10984815B1 (en) Techniques for removing non-linear echo in acoustic echo cancellers
US20190325888A1 (en) Speech recognition method, device, apparatus and computer-readable storage medium
TWI502584B (zh) 電腦實施的波束成形方法，波束成形系統及相關之非暫時性電腦可讀媒體
US11004453B2 (en) Avoiding wake word self-triggering
CN107240396B (zh) 说话人自适应方法、装置、设备及存储介质
TW202027061A (zh) 電子裝置及其語音指令辨識方法
US10529331B2 (en) Suppressing key phrase detection in generated audio using self-trigger detector
TW202008352A (zh) 方位角估計的方法、設備、語音交互系統及儲存介質
CN112236820A (zh) 波束形成器和声学回声消除器(aec)系统
EP3314908A1 (en) Noise reduction for electronic devices
US20210241070A1 (en) Hybrid convolution operation
TW201319927A (zh) 音訊處理系統及音訊暫存器之調整方法
CN108495235B (zh) 一种重低音的分离方法、装置、计算机设备及存储介质
CN112802490B (zh) 一种基于传声器阵列的波束形成方法和装置
CN112669805B (zh) 一种基于方程误差算法的燃气站内压缩机有源噪声控制系统
CN110992975B (zh) 一种语音信号处理方法、装置及终端
WO2020107455A1 (zh) 语音处理方法、装置、存储介质及电子设备
CN104168000A (zh) 音频处理系统及方法
CN111342981A (zh) 局域网环境下设备间的仲裁方法、电子设备、局域网系统
US20220272442A1 (en) Voice processing method, electronic device and readable storage medium
CN115835094A (zh) 音频信号处理方法、系统、设备、产品及介质
CN113194378B (zh) 音频信号的降噪方法、音频信号处理装置及电子设备
CN113205824B (zh) 声音信号处理方法、装置、存储介质、芯片及相关设备
CN111383634B (zh) 根据基于声音的机制停用智能显示设备的显示器的方法及系统

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: 19930399

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19930399

Country of ref document: EP

Kind code of ref document: A1

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 28.03.2022)

122 Ep: pct application non-entry in european phase

Ref document number: 19930399

Country of ref document: EP

Kind code of ref document: A1