Method and system for compensating the frequency response of a microphone

Abstract

The present disclosure provides a method and system for compensating the frequency response of a microphone array. The method includes: multiple microphones in the microphone array respectively receive compensation signals emitted by a calibration speaker and respectively output multiple output signals; based on the multiple output signals, determining the normalized frequency responses of the multiple microphones; calculating a compensation gain for each of the multiple microphones based on the normalized frequency responses; and storing the calculated compensation gain for each microphone.

Description

Technical Field

[0001] The present invention relates to the fields of speech enhancement and speech separation, and in particular to a method and system for compensating the frequency response of a microphone in a vehicle. Background Art

[0002] As vehicles become increasingly intelligent, the fields of in-vehicle speech enhancement and multi-speech separation are receiving increasing attention. Typically, multiple in-vehicle microphones are calibrated to be completely consistent before leaving the factory to ensure the accuracy of the execution results of various speech-related algorithms (e.g., blind source separation algorithms). However, during use, due to various factors such as usage time, temperature, and humidity, the loss of each microphone will vary, resulting in different microphone characteristics. This can lead to many disadvantages. For example, when there is more than one passenger speaking simultaneously in a vehicle, the vehicle's host system often cannot perform speech separation well. This is because the inconsistent degradation of microphone performance means that the predetermined conditions required by the speech separation algorithm are no longer met. This results in a poor user experience for users. In addition, if the damaged microphone needs to be replaced or recalibrated, the user needs to return the car to the factory for repair, which brings inconvenience to the user.

[0003] Therefore, it is necessary to develop a method and system that can compensate for the frequency response of an on-board microphone to improve the accuracy of a speech processing algorithm (eg, blind source separation), thereby providing users with a better user experience. Summary of the Invention

[0004] One or more embodiments of the present invention provide a method for compensating the frequency response of a microphone. The method includes receiving compensation signals emitted by a calibration speaker through multiple microphones in a microphone array and outputting multiple output signals. Based on the multiple output signals, a normalized frequency response of the multiple microphones is determined. Based on the normalized frequency response, a compensation gain is calculated for each microphone in the microphone array; and the calculated compensation gain for each microphone is stored.

[0005] One or more embodiments of the present invention provide a system for compensating the frequency response of a microphone. The system includes a calibration speaker, a microphone array, a processor, and a memory. The calibration speaker is configured to transmit a compensation signal to the microphone array. Multiple microphones in the microphone array receive the compensation signal transmitted by the calibration speaker and output multiple microphone output signals. The processor is configured to: determine a normalized frequency response of the multiple microphones based on the multiple output signals; and calculate a compensation gain for each of the multiple microphones based on the normalized frequency response. The memory is configured to store the calculated compensation gain for each microphone.

[0006] One or more embodiments of the present invention provide a computer-readable medium configured to perform the above-described method steps.

[0007] Advantageously, the method and system for frequency response compensation disclosed in the present invention can conveniently and flexibly improve the accuracy of blind source separation, thereby bringing better user experience to users. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] The system can be better understood with reference to the following description in conjunction with the accompanying drawings. The components in the drawings are not to scale, but rather emphasis is placed on illustrating the principles of the invention. In addition, in the drawings, similar or identical reference numerals represent similar or identical elements.

[0009] Figure 1 is a schematic diagram for exemplifying sound sources and microphones in a vehicle environment.

[0010] Figure 2 is a flow chart of a method for compensating frequency responses of multiple microphones in a microphone array according to one or more embodiments of the present invention.

[0011] Figure 3 FIG. 4 is a schematic diagram of recalibrating microphones in a microphone array so that outputs of the microphones in the microphone array remain consistent according to one or more embodiments of the present invention.

[0012] Figure 4 is a schematic diagram of a microphone array arranged in a circular array or a spherical array according to one or more embodiments of the present invention.

[0013] Figure 5 is a schematic diagram of a microphone array arranged in a linear array according to one or more embodiments of the present invention.

[0014] Figure 6 is a schematic diagram of another microphone array arranged in a linear array according to one or more embodiments of the present invention.

[0015] Figure 7 is a block diagram of a system for compensating frequency responses of multiple microphones in a microphone array according to one or more embodiments of the present invention. DETAILED DESCRIPTION

[0016] It should be understood that the following description of the embodiments is provided for illustrative purposes only and is not intended to be limiting. The division of the examples in the functional blocks, modules, or units shown in the accompanying drawings should not be interpreted as indicating that these functional blocks, modules, or units must be implemented as physically separate units. The functional blocks, modules, or units shown or described may be implemented as separate units, circuits, chips, functions, modules, or circuit elements. One or more functional blocks or units may also be implemented in a common circuit, chip, circuit element, or unit.

[0017] When multiple people speak in a vehicle at the same time, some speech enhancement processing is required during the speech recognition process, such as the BSS algorithm to isolate the clean speech. Figure 1 This is a schematic diagram for illustrating the sound source and microphone in a vehicle environment. For ease of understanding, the figure only uses two microphones as an example and shows only one sound source to illustrate the simple principle of blind source separation. Figure 1 As shown, in a method such as DUET, the signals received at a pair of microphones are represented as N source signals s j (t), j = 1, 2, ..., N, the output signals of the two microphones are denoted as x1(t) and x2(t), which can be expressed as:

[0018]

[0019]

[0020] Where N represents the number of sources, δ j represents the arrival delay between microphones (i.e., the time difference between the sound source and the two microphones), α j is the relative attenuation factor, which corresponds to the attenuation ratio of the path between the source and the microphone.

[0021] The above expressions assume that the microphones have the same frequency response. In other words, when a person speaks into the microphone (j = 1), if the distance between the person and the microphone is the same (α = 1, δ = 0), then the microphone input is the same (s(t)), so the microphone output should be the same, that is, x1(t) = x2(t). However, in use, the properties of the microphone will change with time, temperature, and humidity, and the changes may be inconsistent. Even microphones with the same initial specifications may have inconsistent properties after a period of use. In this case, even though the input is s(t), the output signals x1(t) and x2(t) may be different.

[0022] Figure 2A flow chart of a method for compensating the frequency response of multiple microphones in a microphone array to achieve microphone recalibration according to one or more embodiments is shown. The process of recalibrating the multiple microphones in the microphone array can be initiated by a user or operator based on the speech recognition situation inside the vehicle. For example, the user can trigger the vehicle's microphone calibration start switch or touch screen to send a microphone recalibration control signal to the vehicle host system to initiate the process. Figure 2 The compensation method shown. For example, at S210, after receiving the recalibration control signal, the calibration speaker in the vehicle can emit a compensation signal (i.e., a reference signal). The calibration speaker can be a speaker of a mobile device, a speaker in a vehicle audio system, or any other type of omnidirectional speaker. At S210, multiple microphones in the microphone array receive the compensation signal emitted by the calibration speaker and output multiple microphone output signals respectively. At S220, for example, multiple microphone output signals can be received by a vehicle host system / processor, and based on the multiple output signals, the normalized frequency responses of the multiple microphones are determined. At S230, a compensation gain is calculated for each of the multiple microphones based on the normalized frequency response. At S240, the compensation gain calculated for each microphone can be stored for use when calling a voice processing algorithm (e.g., an algorithm such as BBS).

[0023] Figure 3 The schematic diagram of recalibrating the microphones in the microphone array so that the outputs of the microphones in the microphone array remain consistent is shown as an example. Figure 3 For ease of illustration, the microphone array includes only two microphones. For example, microphone 1 (mic1) and microphone 2 (mic2) can be microphones fixed to the vehicle and coupled to the vehicle's host system (e.g., a processor or controller). A speaker is positioned on the central axis of symmetry of the microphone array. As shown in the figure, distance L is equal to distance R, equidistant from microphones 1 and 2.

[0024] After receiving the recalibration control signal sent by the user to the system, the system is placed in recalibration mode. The control signal controls the speaker to send a compensation signal, microphone 1 and microphone 2 start and record the compensation signal, and output the microphone output signals x1(t) and x2(t) to the vehicle host system respectively. The compensation signal can be a sweep signal (chrip signal) with a broadband frequency and normalized amplitude. For example, the sweep signal can be linearly swept from a frequency of 0.1kHz to a frequency of 4kHz, and its duration is always 5s. For example, the recording duration of the microphone is about 7 However, those skilled in the art will appreciate that the sweep range, duration, and microphone recording time of the sweep signal are provided for illustrative purposes only and are not intended to be limiting. These parameters may be varied according to specific requirements.

[0025] The vehicle host system receives the microphone output signals x1(t) and x2(t) and converts them into frequency-domain signals X1(jω) and X2(jω). The normalized frequency responses (UFRs) of microphones 1 and 2 are then calculated based on the frequency-domain signals. Next, the gains of the two microphones are calculated based on the normalized frequency responses, for example, compensation gain 1 for microphone 1 and gain 2 for microphone 2. Finally, the calculated compensation gains 1 and 2 for the two microphones are stored in the system for use in algorithms such as the BSS algorithm. For example, once the BSS algorithm is invoked, the calibrated and updated microphone gains are first retrieved from memory to compensate for the frequency responses of the microphone output signals. The compensated output signals are then used as input to the BSS algorithm. For example, the spectrum of the audio signal received from microphone 1 is multiplied by gain 1, and the spectrum of the audio signal received from microphone 2 is multiplied by gain 2. Thus, the frequency responses of the output signals of the two microphones are compensated using the corresponding stored compensation gains. This improves the accuracy of subsequent speech processing algorithms (such as the BSS algorithm).

[0026] For the purpose of briefly explaining the principle, Figure 3 Only an example of a microphone array including two microphones is given. However, the microphone array may include more microphones according to actual needs and may have different array arrangements. For example, the microphone array may be a circular array, a spherical array, a linear array, or the like. Figures 4 to 6 The following will refer to Figures 3 to 6 Different microphone arrays are used as examples to illustrate how to compensate for the frequency responses of multiple microphones in different types of microphone arrays.

[0027] Figure 4 A schematic diagram showing that the microphone array is a circular array or a spherical array. Figure 3 and Figure 4 The loudspeakers are all arranged on the central symmetric axis of the microphone array and the distance between the loudspeakers and each microphone is equal.

[0028] for Figure 3 and Figure 4For example, the frequency response of the output signal of any one microphone in the microphone array can be selected as UFR, the compensation gain of the selected microphone can be set to 1, and the compensation gain of each of the remaining microphones can be calculated as the ratio of UFR to the frequency response amplitude of the output signal of each microphone. For example, the gain of the selected microphone can be set to gain = 1, and the gains of the remaining microphones can be calculated as:

[0029] Alternatively, the frequency response amplitudes of the output signals of all or some of the microphones in the microphone array may be calculated separately, and the weighted sum of the frequency response amplitudes may be calculated to thereby calculate the normalized frequency responses of all or some of the microphones.

[0030] UFR=a*|X1(jω)|+b*|X2(jω)|+…+q*|X p (jω)|,

[0031] Where p ≤ N, a + b + ... + q = 1, N represents the total number of microphones in the microphone array, P represents the number of partial microphones in the microphone array, and a, b, ..., p represent the weighting coefficients for the corresponding microphones. For example, the weighting coefficients can be the same, 1 / p, or they can be set based on the importance of the microphones. For example, if a microphone's output is more important, its weighting coefficient will be larger.

[0032] Then, the ratio of UFR to the frequency response amplitude of each microphone output signal can be calculated to calculate the compensation gain of each microphone in the microphone array.

[0033]

[0034] Alternatively, the frequency response energy values ​​of the output signals of all or some of the microphones in the microphone array may be calculated respectively, and the UFRs of all or some of the microphones in the microphone array may be obtained by calculating the weighted frequency response energy values.

[0035] UFR=(a*|X1(jω)| 2 +b*|X2(jω)| 2 +…+q*|X p (jω)| 2 ) 1 / 2 ,

[0036] Wherein, p≤N,a+b+…+q=1, N represents the total number of microphones in the microphone array, P represents the number of some microphones in the microphone array, and a, b…p are weight coefficients of the corresponding microphones respectively.

[0037] Then, the compensation gain of each microphone can be calculated by calculating the ratio of UFR to the frequency response energy of the output signal of each microphone.

[0038]

[0039] Figure 5 and Figure 6 A schematic diagram of a microphone array arranged in a linear array is shown. Figure 5 and Figure 6 In the illustrated linear array, the calibration loudspeaker is placed on the central symmetry axis of the microphone array. In this configuration, the distance between the loudspeaker and each microphone is not equal.

[0040] Figure 5 The microphone array shown in the linear array arrangement includes an even number of microphones. Figure 5 In the illustrated arrangement of an even number of microphones, for example, the number of microphones is N, and the loudspeaker is placed on the central axis of symmetry between the N / 2th and N / 2+1th microphones. To calculate the compensation gain for each microphone, the microphones can be grouped. For example, two microphones at the same distance from the loudspeaker can be grouped together, resulting in a total of N / 2 microphone groups. For example, the first microphone group includes microphones numbered 1 and N, the second microphone group includes microphones numbered 2 and N-1, and so on. The N / 2th microphone group includes microphones numbered N / 2 and N / 2+1.

[0041] For each microphone group, Figure 3 The gain is calculated using the dual-microphone frequency response compensation scheme shown. For example, for the first microphone group, the frequency response of the output signal of microphone number 1 in the group is selected as UFR, the compensation gain of the selected microphone number 1 is set to 1, and the compensation gain of the other microphone numbered N is obtained by calculating the ratio of UFR to the frequency response amplitude of the output signal of another microphone in the group. Thus, for example, the gain of microphone numbered 1 in the group can be set to gain1 = 1, and the gain of the other microphone numbered N in the group can be calculated as:

[0042] This process is deduced in this way until the gains of the microphones numbered N / 2 and N / 2+1 in the N / 2th microphone group are calculated, thereby finally obtaining the compensation gains of all microphones.

[0043] Alternatively, for each microphone group, the frequency response amplitudes of the output signals of the two microphones in the group can be calculated, and the weighted sum of the frequency response amplitudes can be used as the UFR for each group. The compensation gain for each microphone in each microphone group can be obtained by calculating the ratio of the UFR for each group to the frequency response amplitude of the output signal of each microphone in the group.

[0044] For example, for the first group of microphones, the UFR of the group can be calculated using the following formula:

[0045] UFR=a*|X1(jω)|+q*|X N (jω)|,

[0046] Next, the gains of the two microphones in the first microphone group are calculated as:

[0047]

[0048] This process is deduced in this way until the gains of the microphones numbered N / 2 and N / 2+1 in the N / 2th group of microphones are calculated, thereby calculating the gains of all the microphones.

[0049] Alternatively, for each microphone group, the frequency response energy of the output signals of the two microphones in the group can be calculated, and the weighted sum of the frequency response energies can be used as the UFR for each group. The compensation gain for each microphone in the group can then be calculated by calculating the ratio of the UFR for each group to the frequency response energy of the output signal of each microphone in the group.

[0050] For example, for the first group of microphones, the UFR of the group can be calculated using the following formula:

[0051] UFR=(a*|X1(jω)| 2 +q*|X p (jω)| 2 ) 1 / 2 ,

[0052] Then the gains of the two microphones in the first microphone group are calculated as:

[0053]

[0054] This process is deduced in this way until the gains of the microphones numbered N / 2 and N / 2+1 in the N / 2th group of microphones are calculated, thereby calculating the gains of all the microphones.

[0055] Figure 6The microphone array of the linear array arrangement shown includes an odd number of microphones. For example, if the number of microphones is N (N is an odd number), the loudspeaker is placed on the axis of symmetry of the (N+1) / 2th microphone array. In addition to the (N+1) / 2th microphone, two microphones with the same distance are grouped together, resulting in a total of (N-1) / 2 microphone groups. For example, the first group of microphones includes two microphones numbered 1 and N, the second group of microphones includes microphones numbered 2 and N-1, ..., the (N-1) / 2 group of microphones includes microphones numbered (N-1) / 2 and (N+1) / 2+1. In this case, the gain of the (N+1) / 2th microphone is individually set to gain((N+1) / 2)=1. For other groups of microphones, the reference Figure 5 The same method is used to calculate the gain of each microphone, thereby finally obtaining the gains of all microphones gain1, ..., gain(N-1) / 2, gain((N+1) / 2+1), ..., gain N.

[0056] Figure 7 FIG. 1 shows a block diagram of a system for compensating the frequency response of a microphone according to various embodiments of the present invention. Figure 7 As shown, the system includes a calibration speaker 701, a microphone array 702, a processor 703, and a memory 704. When the system is in microphone recalibration mode, the calibration speaker 701 emits a compensation signal. The multiple microphones in the microphone array 702 receive the compensation signal from the calibration speaker 701 and output multiple microphone output signals to the processor 703. Based on the multiple output signals output by the microphone array 702, the processor 703 determines the normalized frequency responses of the multiple microphones. Based on the normalized frequency responses, it calculates a compensation gain for each microphone in the microphone array and stores the calculated compensation gain in the memory 704.

[0057] Processor 703 is further configured to determine whether the calibration speaker is positioned at an equal distance from each microphone in the microphone array. When the processor determines that the calibration speaker is positioned at an equal distance from each microphone in the microphone array, the processor may select a frequency response of an output signal of one microphone in the microphone array as a normalized frequency response UFR, set a compensation gain for the selected microphone to 1, and calculate a compensation gain for each of the remaining microphones in the microphone array as a ratio of the normalized frequency response UFR to a frequency response amplitude of the output signal of each microphone.

[0058] Furthermore, the processor 703 is also configured to, when it is determined that the calibration speaker is positioned at an equal distance from each microphone in the microphone array, calculate the frequency response amplitudes of all or part of the multiple output signals and use the weighted sum of the frequency response amplitudes as the normalized frequency response UFR, and calculate the compensation gain of each microphone in the microphone array as the ratio of the normalized frequency response UFR to the frequency response amplitude of the output signal of each microphone.

[0059] Furthermore, the processor 703 is further configured to, when it is determined that the calibration speaker is positioned at an equal distance from each microphone in the microphone array, calculate the frequency response energy of all or part of the multiple output signals and use the weighted sum of the frequency response energies as the normalized frequency response UFR, and set the compensation gain in each of the multiple microphones to the ratio of the normalized frequency response UFR to the frequency response energy of the output signal of each microphone.

[0060] Furthermore, the processor 703 is further configured to, when the processor determines that the calibration speaker is not positioned at an equal distance from each microphone in the microphone array, further determine whether the calibration speaker is located on the central symmetry axis of the microphone array. If the calibration speaker is located on the central symmetry axis of the microphone array and the number of microphones in the microphone array is an even number, the plurality of microphones are grouped by grouping two microphones that are at the same distance from the calibration speaker. If the calibration speaker is located on the central symmetry axis of the microphone array and the number of microphones in the microphone array is an odd number, the plurality of microphones other than the microphones located on the central symmetry axis are grouped by grouping two microphones that are at the same distance from the calibration speaker.

[0061] Furthermore, the processor 703 is further configured to, when the number of microphones in the microphone array is an even number, select the frequency response of the output signal of one microphone in each group of microphones as the normalized frequency response of each group, set the compensation gain of the selected one microphone to 1, and calculate the compensation gain of the other microphone in each group of microphones as the ratio of the normalized frequency response of each group to the frequency response amplitude of the output signal of the other microphone in each group of microphones.

[0062] Furthermore, the processor 703 is further configured to, when the number of microphones in the microphone array is an even number, calculate, for each group of microphones, the frequency response amplitude of the output signal of each group of microphones and use the weighted sum of the frequency response amplitudes of the output signals as the normalized frequency response of each group, and calculate the compensation gain of each microphone in each group of microphones as the ratio of the normalized frequency response of each group to the frequency response amplitude of the output signal of each microphone.

[0063] Furthermore, the processor 703 is further configured to, when the processor determines that the number of microphones in the microphone array is an even number, calculate the frequency response energy of multiple output signals of each group of microphones for each group of microphones and use the weighted sum of the frequency response energies as the normalized frequency response of each group, and calculate the compensation gain of each microphone in each group of microphones as the ratio of the normalized frequency response of each group to the frequency response energy of each microphone output signal.

[0064] Furthermore, the processor 703 is further configured to, when the number of microphones in the microphone array is an odd number, select the frequency response of the output signal of one microphone in each group of microphones as the normalized frequency response of each group, set the compensation gain of the selected one microphone to 1, calculate the compensation gain of another microphone in each group of microphones as the ratio of the normalized frequency response of each group to the frequency response amplitude of the output signal of another microphone in each group of microphones, and set the compensation gain of the microphone located on the central symmetry axis to 1.

[0065] Furthermore, the processor 703 is further configured to, when the number of microphones in the microphone array is an odd number, calculate, for each group of microphones, the frequency response amplitude of the output signal of each group of microphones and use the weighted sum of the frequency response amplitudes of the output signals as the normalized frequency response of each group, calculate the compensation gain of each microphone in each group of microphones as the ratio of the normalized frequency response of each group to the frequency response amplitude of the output signal of each microphone, and set the compensation gain of the microphone located on the central symmetry axis to 1.

[0066] Furthermore, the processor 703 is further configured to, when the number of microphones in the microphone array is an odd number, calculate, for each group of microphones, the frequency response energy of the output signal of each group of microphones and use the weighted sum of the frequency response energies of the output signals as the normalized frequency response of each group, calculate the compensation gain of each microphone in each group of microphones as the ratio of the normalized frequency response of each group to the frequency response energy of the output signal of each microphone, and set the compensation gain of the microphone located on the central symmetry axis to 1.

[0067] The processor of the present invention as a whole may be a microprocessor, an application-specific integrated circuit (ASIC), a system on a chip (SoC), a mobile computing device (eg, a tablet computer or a mobile phone), a media player, etc.

[0068] Any one or more of the processors, memories, or systems described herein include computer-executable instructions that can be compiled or interpreted from a computer program created using various programming languages ​​and / or techniques. In general, a processor (such as a microprocessor) receives instructions from, for example, a memory, a computer-readable medium, or the like and executes the instructions. The processor includes a non-transitory computer-readable storage medium capable of executing the instructions of the software program. The computer-readable medium can be, but is not limited to, an electronic storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination thereof.

[0069] The description of the embodiments has been presented for the purpose of illustration and description. Suitable modifications and variations of the embodiments may be performed in view of the above description or may be obtained by practical methods. For example, unless otherwise noted, the one or more methods described may be performed by a combination of suitable devices and / or systems. The method may be performed in the following manner: one or more logic devices (e.g., processors) may be used in conjunction with one or more other hardware elements (such as storage devices, memories, circuits, hardware network interfaces, etc.) to execute stored instructions. The method and associated actions may also be performed in parallel and / or simultaneously in various sequences other than the sequence described in this application. The system is exemplary in nature and may include additional elements and / or omit elements. The subject matter of the present disclosure includes all novel and non-obvious combinations of the disclosed various methods and system configurations and other features, functions and / or properties.

[0070] As used in this application, elements or steps listed in the singular and preceded by the word "one" should be understood as not excluding a plurality of said elements or steps, unless such exclusion is indicated. In addition, reference to "one embodiment" or "an example" of the present disclosure is not intended to be interpreted as excluding the existence of another embodiment that also incorporates the enumerated features. The present invention has been described above with reference to specific embodiments. However, it will be understood by those skilled in the art that various modifications and changes may be made thereto without departing from the broader spirit and scope of the present invention as set forth in the appended claims.

Claims

1. A method for compensating the frequency response of a microphone, comprising: The plurality of microphones in the microphone array receive the compensation signal emitted by the calibration speaker and output a plurality of output signals; determining normalized frequency responses of the plurality of microphones based on the plurality of output signals; calculating a compensation gain for each of the plurality of microphones based on the normalized frequency response; storing the calculated compensation gain for each microphone; as well as Determining whether the calibration speaker is positioned at an equal distance from each microphone in the microphone array, wherein: When the calibration speaker is not positioned at an equal distance from each microphone in the microphone array, determining whether the calibration speaker is located on the central symmetry axis of the microphone array, If the calibration speaker is located on the central symmetry axis of the microphone array and the number of microphones in the microphone array is even, the plurality of microphones are grouped by grouping two microphones that are at the same distance from the calibration speaker into one group, If the calibration speaker is located on the central symmetric axis of the microphone array and the number of microphones in the microphone array is an odd number, grouping the plurality of microphones except for the microphone located on the central symmetric axis by grouping two microphones at the same distance from the calibration speaker into one group, and When the number of microphones in the microphone array is an even number, a frequency response of an output signal of one microphone in each group of microphones is selected as a normalized frequency response of each group, a compensation gain of the selected one microphone is set to 1, and a compensation gain of another microphone in each group of microphones is calculated as a ratio of the normalized frequency response of each group to a frequency response amplitude of an output signal of another microphone in each group of microphones.

2. The method according to claim 1, wherein When the calibration speaker is positioned at an equal distance from each microphone in the microphone array, Selecting a frequency response of an output signal of a microphone in the microphone array as the normalized frequency response, Set the makeup gain of the selected microphone to 1, and The compensation gain of each of the remaining microphones of the microphone array is calculated as a ratio of the normalized frequency response to a frequency response amplitude of an output signal of each microphone.

3. The method according to claim 1, wherein When the calibration speaker is positioned at an equal distance from each microphone in the microphone array, the frequency response amplitudes of all or part of the multiple output signals are calculated and a weighted sum of the frequency response amplitudes is used as the normalized frequency response, and the compensation gain of each microphone in the microphone array is calculated as the ratio of the normalized frequency response to the frequency response amplitude of the output signal of each microphone.

4. The method according to claim 1, wherein When the calibration speaker is positioned at an equal distance from each microphone in the microphone array, frequency response energy of all or part of the multiple output signals is calculated and a weighted sum of the frequency response energies is used as the normalized frequency response, and a compensation gain in each of the multiple microphones is calculated as a ratio of the normalized frequency response to the frequency response energy of the output signal of each microphone.

5. The method according to claim 1, wherein When the number of microphones in the microphone array is an even number, for each group of microphones, the frequency response amplitude of the output signals of the microphones in each group is calculated and a weighted sum of the frequency response amplitudes of the output signals is used as the normalized frequency response of each group, and the compensation gain of each microphone in each group of microphones is calculated as the ratio of the normalized frequency response of each group to the frequency response amplitude of the output signal of each microphone.

6. The method according to claim 1, wherein When the number of microphones in the microphone array is an even number, for each group of microphones, frequency response energies of multiple output signals of each group of microphones are calculated and a weighted sum of the frequency response energies is used as a normalized frequency response of each group, and a compensation gain of each microphone in each group of microphones is calculated as a ratio of the normalized frequency response of each group to the frequency response energy of the output signal of each microphone.

7. The method according to claim 1, wherein When the number of microphones in the microphone array is an odd number, a frequency response of an output signal of one microphone in each group of microphones is selected as a normalized frequency response of each group, a compensation gain of the selected one microphone is set to 1, a compensation gain of another microphone in each group of microphones is calculated as a ratio of the normalized frequency response of each group to a frequency response amplitude of an output signal of another microphone in each group of microphones, and the compensation gain of the microphone located on the central symmetry axis is set to 1.

8. The method according to claim 1, wherein When the number of microphones in the microphone array is an odd number, for each group of microphones, the frequency response amplitude of the output signal of the microphones of each group is calculated and the weighted sum of the frequency response amplitudes of the output signals is used as the normalized frequency response of each group, the compensation gain of each microphone in each group of microphones is calculated as the ratio of the normalized frequency response of each group to the frequency response amplitude of the output signal of each microphone, and the compensation gain of the microphone located on the central symmetry axis is set to 1.

9. The method according to claim 1, wherein When the number of microphones in the microphone array is an odd number, for each group of microphones, the frequency response energy of the output signals of the microphones of each group is calculated and the weighted sum of the frequency response energies of the output signals is used as the normalized frequency response of each group, the compensation gain of each microphone in each group of microphones is calculated as the ratio of the normalized frequency response of each group to the frequency response energy of the output signal of each microphone, and the compensation gain of the microphone located on the central symmetry axis is set to 1.

10. A system for compensating the frequency response of a microphone, comprising: a calibration speaker configured to emit a compensation signal; a microphone array comprising a plurality of microphones, wherein the plurality of microphones of the microphone array receive the compensation signal emitted by the calibration speaker and output a plurality of microphone output signals; A processor configured to: determining normalized frequency responses of the plurality of microphones based on the plurality of output signals; calculating a compensation gain for each of the plurality of microphones based on the normalized frequency response; Determining whether the calibration speaker is positioned at an equal distance from each microphone in the microphone array, wherein: When the calibration speaker is not positioned at an equal distance from each microphone in the microphone array, determining whether the calibration speaker is located on the central symmetry axis of the microphone array, If the calibration speaker is located on the central symmetry axis of the microphone array and the number of microphones in the microphone array is even, the plurality of microphones are grouped by grouping two microphones that are at the same distance from the calibration speaker into one group, If the calibration speaker is located on the central symmetric axis of the microphone array and the number of microphones in the microphone array is an odd number, grouping the plurality of microphones except for the microphone located on the central symmetric axis by grouping two microphones at the same distance from the calibration speaker into one group, and When the number of microphones in the microphone array is an even number, selecting a frequency response of an output signal of one microphone in each group of microphones as a normalized frequency response of each group, setting a compensation gain of the selected one microphone to 1, and calculating a compensation gain of another microphone in each group of microphones as a ratio of the normalized frequency response of each group to a frequency response amplitude of an output signal of another microphone in each group of microphones; and A memory is configured to store the calculated compensation gain for each microphone.

11. A computer-readable medium comprising instructions for executing the method of any one of claims 1 to 9.

Images