Speech enhancement method and device, equipment, storage medium and program product

Abstract

The invention provides a speech enhancement method and device, equipment, a storage medium and a program product. The method comprises the following steps: acquiring a to-be-processed voice signal; wherein the to-be-processed voice signal represents voice signals collected by a plurality of microphones; determining a signal processing mode matched with the to-be-processed voice signal based on the to-be-processed voice signal; wherein the signal processing mode represents a processing mode for performing enhancement processing on the voice signal to be processed; and according to the matched signal processing mode, performing noise reduction processing on the to-be-processed voice signal to obtain a voice signal enhancement result. According to the method provided by the invention, the voice enhancement method can be suitable for various noise scenes by combining various microphones, various voice communication requirements can be met, and the voice communication quality is ensured.

Description

Technical Field

[0001] The present application relates to the field of communication technology, and in particular to a speech enhancement method, apparatus, device, storage medium and program product. Background Art

[0002] Headphones have become a common device in daily life due to their convenience and ease of use. How to ensure the voice communication quality of headphones in various scenarios has become an urgent problem to be solved.

[0003] At present, in high-noise scenarios, air conduction microphone arrays or physical isolation methods are mainly used to achieve speech noise reduction (also known as speech enhancement) to meet the voice communication requirements of headphones.

[0004] However, in some scenarios, for example, when a protective mask or face shield is required, the air conduction microphone cannot be used for voice communication, which affects the voice communication quality of the headset. In addition, the above-mentioned voice noise reduction method has high requirements for the structural design of the headset (or wearable device), making the structural design of the headset (or wearable device) highly challenging. Summary of the invention

[0005] The present application provides a speech enhancement method, apparatus, device, storage medium and program product to solve the problem of poor voice communication quality in high noise scenes.

[0006] In a first aspect, the present application provides a speech enhancement method, comprising:

[0007] Acquire a speech signal to be processed; wherein the speech signal to be processed represents speech signals collected by multiple microphones;

[0008] Based on the speech signal to be processed, determining a signal processing mode that matches the speech signal to be processed; wherein the signal processing mode represents a processing method for enhancing the speech signal to be processed;

[0009] According to the matched signal processing mode, the speech signal to be processed is subjected to noise reduction processing to obtain a speech signal enhancement result.

[0010] In one example, the multiple microphones include a first microphone for collecting voice signals transmitted through the air, a second microphone for collecting voice signals transmitted through bones, and a third microphone for collecting noise in the air; and determining a signal processing mode matching the voice signal to be processed based on the voice signal to be processed includes:

[0011] Based on the first voice signal collected by the second microphone, voice activity detection is performed to obtain a voice activity detection result; wherein the voice activity detection result indicates whether the second microphone collects the voice signal of the current speaker;

[0012] Based on the voice activity detection result, a noise signal is determined from the second voice signal collected by the first microphone, and a first signal-to-noise ratio of the second voice signal is determined; wherein the noise signal represents the noise of the collection environment where the signal to be processed is located;

[0013] Based on the first signal-to-noise ratio, a signal processing mode matching the speech signal to be processed is determined.

[0014] In one example, determining a signal processing mode matching the speech signal to be processed based on the first signal-to-noise ratio includes:

[0015] When it is determined that the first signal-to-noise ratio is greater than a preset first value, determining that a signal processing mode matching the speech signal to be processed is a first processing mode;

[0016] Among them, the first processing mode represents a processing method in which, based on the voice activity detection result corresponding to the first voice signal, a noise signal is determined from the second voice signal, and the second voice signal is subjected to noise reduction processing based on the noise signal to obtain a voice signal enhancement result.

[0017] In one example, determining a signal processing mode matching the speech signal to be processed based on the first signal-to-noise ratio includes:

[0018] When it is determined that the first signal-to-noise ratio is less than the preset first value and greater than a preset second value, determining a second signal-to-noise ratio corresponding to the first speech signal;

[0019] In the case where it is determined that the second signal-to-noise ratio is greater than a preset third value, determining that the signal processing mode matching the speech signal to be processed is the second processing mode;

[0020] The second processing mode represents a processing method for obtaining a speech signal enhancement result based on a low-frequency speech signal determined based on the first speech signal and a high-frequency speech signal determined based on the second speech signal.

[0021] In one example, the low-frequency speech signal determined based on the first speech signal and the high-frequency speech signal determined based on the second speech signal are used to obtain a speech signal enhancement result, including:

[0022] Based on a preset low-pass filter, the first voice signal and the third voice signal collected by the third microphone are filtered, and based on the filtered third voice signal, the filtered first voice signal is subjected to noise reduction processing to obtain the low-frequency voice signal; wherein the third voice signal represents the noise of the collection environment where the voice signal to be processed is located;

[0023] Based on the low-frequency voice signal, a high-pass filter is set; wherein the high-pass filter is determined based on the signal bandwidth of the low-frequency voice signal;

[0024] Based on the high-pass filter, the second voice signal and the third voice signal are filtered, and based on the filtered third voice signal, the filtered second voice signal is denoised to obtain the high-frequency voice signal;

[0025] The low-frequency speech signal and the high-frequency speech signal are fused, and the fused speech signal is subjected to noise reduction processing to obtain the speech signal enhancement result.

[0026] In one example, determining a signal processing mode matching the speech signal to be processed based on the first signal-to-noise ratio includes:

[0027] When it is determined that the first signal-to-noise ratio is less than the preset second value, or when it is determined that the second signal-to-noise ratio is less than the preset third value, determining that the signal processing mode matching the speech signal to be processed is the third processing mode;

[0028] The third processing mode represents a processing method of performing noise reduction processing on the first voice signal based on the third voice signal collected by the third microphone to obtain a voice signal enhancement result.

[0029] In one example, the noise reduction processing is performed on the speech signal to be processed according to the matched signal processing mode to obtain a speech signal enhancement result, including:

[0030] Acquire an adaptive filter; wherein the adaptive filter is used to control the signal bandwidth information of the speech signal to be processed during the noise reduction process;

[0031] Based on the adaptive filter and the matching signal processing mode, the speech signal to be processed is subjected to noise reduction processing to obtain the speech signal enhancement result.

[0032] In a second aspect, the present application provides a speech enhancement device, comprising:

[0033] An acquisition unit, configured to acquire a speech signal to be processed; wherein the speech signal to be processed represents speech signals collected by a plurality of microphones;

[0034] A determination unit, configured to determine, based on the speech signal to be processed, a signal processing mode that matches the speech signal to be processed; wherein the signal processing mode represents a processing method for enhancing the speech signal to be processed;

[0035] The processing unit is used to perform noise reduction processing on the speech signal to be processed according to the matched signal processing mode to obtain a speech signal enhancement result.

[0036] In one example, the plurality of microphones include a first microphone for collecting voice signals transmitted through air, a second microphone for collecting voice signals transmitted through bones, and a third microphone for collecting noise in the air.

[0037] In one example, the determining unit is used to:

[0038] Based on the first voice signal collected by the second microphone, voice activity detection is performed to obtain a voice activity detection result; wherein the voice activity detection result indicates whether the second microphone collects the voice signal of the current speaker;

[0039] Based on the voice activity detection result, a noise signal is determined from the second voice signal collected by the first microphone, and a first signal-to-noise ratio of the second voice signal is determined; wherein the noise signal represents the noise of the collection environment where the signal to be processed is located;

[0040] Based on the first signal-to-noise ratio, a signal processing mode matching the speech signal to be processed is determined.

[0041] In one example, the determining unit includes:

[0042] A first sub-determining module is used to determine that the signal processing mode matching the speech signal to be processed is a first processing mode when it is determined that the first signal-to-noise ratio is greater than a preset first value;

[0043] Among them, the first processing mode represents a processing method in which, based on the voice activity detection result corresponding to the first voice signal, a noise signal is determined from the second voice signal, and the second voice signal is subjected to noise reduction processing based on the noise signal to obtain a voice signal enhancement result.

[0044] In one example, the determining unit includes:

[0045] A second sub-determination module is used to determine a second signal-to-noise ratio corresponding to the first speech signal when it is determined that the first signal-to-noise ratio is less than the preset first value and greater than a preset second value;

[0046] In the case where it is determined that the second signal-to-noise ratio is greater than a preset third value, determining that the signal processing mode matching the speech signal to be processed is the second processing mode;

[0047] The second processing mode represents a processing method for obtaining a speech signal enhancement result based on a low-frequency speech signal determined based on the first speech signal and a high-frequency speech signal determined based on the second speech signal.

[0048] In one example, the second sub-determining module is used to:

[0049] Based on a preset low-pass filter, the first voice signal and the third voice signal collected by the third microphone are filtered, and based on the filtered third voice signal, the filtered first voice signal is subjected to noise reduction processing to obtain the low-frequency voice signal; wherein the third voice signal represents the noise of the collection environment where the voice signal to be processed is located;

[0050] Based on the low-frequency voice signal, a high-pass filter is set; wherein the high-pass filter is determined based on the signal bandwidth of the low-frequency voice signal;

[0051] Based on the high-pass filter, the second voice signal and the third voice signal are filtered, and based on the filtered third voice signal, the filtered second voice signal is denoised to obtain the high-frequency voice signal;

[0052] The low-frequency speech signal and the high-frequency speech signal are fused, and the fused speech signal is subjected to noise reduction processing to obtain the speech signal enhancement result.

[0053] In one example, the determining unit includes:

[0054] a third sub-determining module, configured to determine that the signal processing mode matching the speech signal to be processed is a third processing mode when it is determined that the first signal-to-noise ratio is less than the preset second value, or when it is determined that the second signal-to-noise ratio is less than the preset third value;

[0055] The third processing mode represents a processing method of performing noise reduction processing on the first voice signal based on the third voice signal collected by the third microphone to obtain a voice signal enhancement result.

[0056] In one example, the processing unit is used to:

[0057] Acquire an adaptive filter; wherein the adaptive filter is used to control the signal bandwidth information of the speech signal to be processed during the noise reduction process;

[0058] Based on the adaptive filter and the matching signal processing mode, the speech signal to be processed is subjected to noise reduction processing to obtain the speech signal enhancement result.

[0059] In a third aspect, the present application provides a computer device, comprising: a processor, and a memory communicatively connected to the processor;

[0060] The memory stores computer-executable instructions;

[0061] The processor executes the computer-executable instructions stored in the memory to implement the method described in the first aspect.

[0062] In a fourth aspect, the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer-executable instructions, and when the computer-executable instructions are executed by a processor, they are used to implement the method described in the first aspect.

[0063] In a fifth aspect, the present application provides a computer program product, comprising computer execution instructions, which are stored in a readable storage medium. At least one processor of a computer device can read the computer execution instructions from the readable storage medium, and the at least one processor executes the computer execution instructions, so that the computer device executes the method described in the first aspect.

[0064] The speech enhancement method, device, equipment, storage medium and program product provided by the present application can collect the speech signal to be processed through a variety of microphones. Since different microphones can collect different signals, at this time, the signal processing mode matching the speech signal to be processed can be directly determined based on the different signals collected by different microphones, which reduces the complexity of signal calculation and improves the calculation efficiency. At the same time, by determining the signal processing mode matching the signal to be processed, the matching speech enhancement method can be adaptively determined according to the acquisition environment of the signal to be processed, so that the speech enhancement method is more diverse, avoiding the problem of poor performance of the speech enhancement method caused by using a single speech enhancement method. Afterwards, based on the matching signal processing mode, the signal to be processed is subjected to noise reduction processing to obtain the speech signal enhancement result, which can ensure the accuracy of the speech signal enhancement result, thereby ensuring the quality of voice communication. At the same time, when this speech enhancement method is applied in headphones, different signal processing modes can be used to reduce the power consumption of the headphones as much as possible, extend the working time of the headphones, and improve the user's experience of using the headphones. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0066] Figure 1 A flowchart of a speech enhancement method provided in an embodiment of the present application;

[0067] Figure 2 A flowchart of another speech enhancement method provided in an embodiment of the present application;

[0068] Figure 3 A schematic diagram of multiple microphones in a headset provided in an embodiment of the present application;

[0069] Figure 4 A flowchart of a speech enhancement method provided in an embodiment of the present application in an actual application scenario;

[0070] Figure 5 A schematic diagram of an application flow of a speech enhancement method provided in an embodiment of the present application;

[0071] Figure 6 A schematic diagram of the structure of a speech enhancement device provided in an embodiment of the present application;

[0072] Figure 7 A schematic diagram of the structure of another speech enhancement device provided in an embodiment of the present application;

[0073] Figure 8 A schematic diagram of the structure of a computer device provided in an embodiment of the present application;

[0074] Fig. 9 The invention is a block diagram of a computer device according to an exemplary embodiment.

[0075] The above drawings have shown clear embodiments of the present application, which will be described in more detail later. These drawings and text descriptions are not intended to limit the scope of the present application in any way, but to illustrate the concept of the present application to those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION

[0076] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0077] The term "and / or" herein only describes an association relationship, indicating that three relationships may exist. For example, A and / or B may represent the following three situations: A exists alone, A and B exist at the same time, and B exists alone. In addition, the term "at least one" herein represents any combination of at least two of any one or more of a plurality of. For example, including at least one of A, B, and C may represent including any one or more elements selected from the set consisting of A, B, and C.

[0078] Due to its convenience and ease, earphones have become a common device in daily life. How to ensure the voice communication quality of earphones in various scenarios (for example, high noise scenarios) has become an urgent problem to be solved.

[0079] Headphones can use air conduction microphones or bone conduction microphones for voice communication. Bone conduction microphones mainly collect voice signals generated by contact and vibration with the speaker's vocal cords, skull, etc. to achieve voice communication; air conduction microphones mainly collect voice signals based on air propagation to achieve voice communication.

[0080] Among them, bone conduction microphones can shield environmental noise from the sound source and obtain relatively pure voice signals. Moreover, due to the low-pass characteristics of the human body, bone conduction microphones cannot collect high-frequency voice signals. Therefore, bone conduction microphones have important application value in complex noise environments.

[0081] However, due to the characteristics of the voice signal collected by the bone conduction microphone, the low-frequency component is heavy and the high-frequency attenuation is serious, the high-frequency information of the voice signal collected by the bone conduction microphone is lost, some unvoiced syllables are missing, the sound is dull and not clear enough, resulting in low intelligibility of the voice signal. Moreover, the clarity of bone conduction speech is related to the position of the bone conduction microphone and the individual's physical characteristics, which makes it impossible to reconstruct the voice signal collected by the bone conduction microphone using a set of parameters. These reasons limit the use scenarios of bone conduction microphones.

[0082] Therefore, it is necessary to improve the quality of voice communication and speech intelligibility when using bone conduction microphones.

[0083] With the development of deep learning, deep neural networks can be used in related technologies to enhance the speech of bone conduction microphones. However, due to the computing power and power consumption limitations of Bluetooth headsets or wearable devices, headsets or devices using bone conduction microphones have not yet been put into practical use.

[0084] Therefore, for headphones, which are commonly used equipment in daily work, it is necessary to improve the adaptability of headphones to various noise environments and extend the working time of headphones.

[0085] Based on this, in the related technology, air conduction microphones are mainly used to collect voice signals to achieve voice communication. In one example, in a high-noise scenario, an air conduction microphone array or a physical isolation method is generally used to achieve noise reduction of voice signals to meet voice communication requirements. This places high demands on the structural design of headphones (e.g., Bluetooth headphones) or wearable devices, and also adds many restrictions to the use methods or usage scenarios of headphones (e.g., Bluetooth headphones) or wearable devices.

[0086] In another example, in a scenario where a protective isolation mask or face mask is required, the wearing of the protective isolation mask or face mask causes the voice signal to be unable to propagate through the air to the air conduction microphone, and thus the air conduction microphone cannot be used for voice communication.

[0087] The speech enhancement method provided in this application is intended to solve the above technical problems in the prior art.

[0088] The technical solution of the present application and how the technical solution of the present application solves the above-mentioned technical problems are described in detail below with specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present application will be described below in conjunction with the accompanying drawings.

[0089] Figure 1 A flow chart of a speech enhancement method provided in an embodiment of the present application is shown as follows: Figure 1 As shown, the method includes:

[0090] S101: Obtain a voice signal to be processed.

[0091] The speech signal to be processed represents the speech signals collected by various microphones.

[0092] In one example, the voice signal to be processed can be collected by a microphone. In this case, the voice signal to be processed may include a pure voice signal and / or a noise signal, wherein the pure voice signal indicates the collected voice signal of the current speaker, and the noise signal indicates a signal other than the voice signal of the current speaker that can be transmitted through the air. For example, the noise signal may be the voice signal of other speakers in the collection environment where the voice signal to be processed is located, or other sound signals in the collection environment where the voice signal to be processed is located, such as wind, rain, whistle, etc. In this case, the type of the noise signal may be one or more, which is not specifically limited here.

[0093] In one example, multiple microphones used to collect speech signals to be processed can focus on collecting different types of signals according to the type of microphone and the installation position of the microphone. For example, when the multiple microphones can represent bone conduction microphones and air conduction microphones, the air conduction microphone mainly focuses on collecting signals based on air propagation. At this time, the air-propagation signal can include the speaker's speech signal and noise signal; the bone conduction microphone is mainly used to collect the current speaker's speech signal transmitted through the bones.

[0094] In one example, the number of each type of microphone is at least one, for example, the number of air conduction microphones may be 1, and the number of bone conduction microphones may be 2. In the case where the number of bone conduction microphones is 2, one of the bone conduction microphones may be installed inside the earphone and in contact with the skin of the current earphone user, and is mainly used to collect the voice signal of the current speaker. In this case, the bone conduction microphone may be recorded as a bone conduction main microphone. The other bone conduction microphone may be installed outside the earphone, and is not in contact with the skin of the current earphone user, and is mainly used to collect sound signals (i.e., noise signals) transmitted through the air. In this case, the bone conduction microphone may be recorded as a bone conduction reference microphone.

[0095] S102: Based on the speech signal to be processed, determine a signal processing mode that matches the speech signal to be processed.

[0096] The signal processing mode represents the processing method for enhancing the speech signal to be processed.

[0097] In one example, the noise environment of the collection environment of the speech signal to be processed can be determined based on the speech signal to be processed, so that different signal processing modes can be determined for different noise environments, so as to perform speech enhancement processing on the speech signal to be processed based on different signal processing modes. The noise environment can characterize the size of the noise signal in the speech signal to be processed collected by the microphone.

[0098] In one example, the signal processing mode can be used to determine a microphone for collecting a speech signal to be processed, and to determine a processing method / processing flow for processing the speech signal to be processed.

[0099] S103: Perform noise reduction processing on the speech signal to be processed according to the matched signal processing mode to obtain a speech signal enhancement result.

[0100] In one example, the acquired speech signal to be processed may be subjected to noise reduction processing according to the processing method / processing flow determined by the matching signal processing mode to achieve speech enhancement and obtain a speech signal enhancement result.

[0101] It can be seen from the above description that the embodiment of the present application can collect the voice signal to be processed through a variety of microphones. Since different microphones can collect different signals, at this time, the signal processing mode matching the voice signal to be processed can be determined directly based on the different signals collected by different microphones, which reduces the complexity of signal calculation and improves the calculation efficiency. At the same time, by determining the signal processing mode matching the signal to be processed, the matching voice enhancement method can be adaptively determined according to the acquisition environment of the signal to be processed, so that the voice enhancement method is more diverse, avoiding the problem of poor performance of the voice enhancement method caused by using a single voice enhancement method. Afterwards, based on the matching signal processing mode, the signal to be processed is subjected to noise reduction processing to obtain the voice signal enhancement result, which can ensure the accuracy of the voice signal enhancement result, thereby ensuring the quality of voice communication. At the same time, when this voice enhancement method is applied in headphones, different signal processing modes can be used to reduce the power consumption of the headphones as much as possible, extend the working time of the headphones, and improve the user's experience of using the headphones.

[0102] Figure 2 A flow chart of another speech enhancement method provided in an embodiment of the present application is shown as follows: Figure 2 As shown, the method includes:

[0103] S201: Obtain a voice signal to be processed.

[0104] The speech signal to be processed represents the speech signals collected by various microphones.

[0105] In an example, this step can refer to the content described in S101 above, and will not be described in detail here.

[0106] In one example, the plurality of microphones may include a first microphone for collecting voice signals transmitted through the air, a second microphone for collecting voice signals transmitted through bones, and a third microphone for collecting noise in the air.

[0107] In an example, Figure 3 The headset shown may include multiple microphones, wherein the first microphone may be understood as an air conduction microphone, the second microphone may be understood as a bone conduction main microphone, and the third microphone may be understood as a bone conduction reference microphone. Figure 3 The bone conduction vibrator in the receiver is a component used to play downlink data.

[0108] S202: Perform voice activity detection on the first voice signal collected by the second microphone to obtain a voice activity detection result.

[0109] In one example, the voice activity detection result may indicate whether the second microphone has collected the voice signal of the current speaker. In this case, the first voice signal may be detected by a voice activity detection algorithm (Voice Activity Detection, VAD) to obtain the voice activity detection result.

[0110] S203: Based on the voice activity detection result, determine a noise signal from the second voice signal collected by the first microphone, and determine a first signal-to-noise ratio of the second voice signal.

[0111] The noise signal represents the noise in the collection environment of the speech signal to be processed, that is, it represents the sound signals in the collection environment of the speech signal to be processed other than the speech signal of the current speaker, such as wind, rain, whistle, etc.

[0112] In one example, when it is determined based on the voice activity detection result that the voice signal of the current speaker is not collected (that is, the current speaker is in a silent state), the second voice signal when the current speaker is in a silent state can be determined as a noise signal.

[0113] Then, a first signal-to-noise ratio of the second speech signal may be calculated based on the second speech signal when the current speaker is not in a silent state and the noise signal.

[0114] S204: Determine a signal processing mode that matches the speech signal to be processed based on the first signal-to-noise ratio.

[0115] The signal processing mode represents the processing method for enhancing the speech signal to be processed.

[0116] In the above implementation, the first signal-to-noise ratio of the second speech signal collected by the air conduction microphone can be calculated to conveniently and quickly determine the noise of the current collection environment, so as to more accurately determine the signal processing mode that matches the speech signal to be processed. The speech enhancement method can be adaptively adjusted according to the actual noise environment, thereby improving the flexibility and diversity of the speech enhancement method, reducing the calculation amount of speech enhancement, and saving power consumption.

[0117] Wherein, based on the first signal-to-noise ratio, determining the signal processing mode that matches the speech signal to be processed can be divided into the following three cases.

[0118] Case 1: When it is determined that the first signal-to-noise ratio is greater than a preset first value, the signal processing mode matching the speech signal to be processed is determined to be the first processing mode.

[0119] Among them, the first processing mode represents a processing method in which a noise signal is determined from a second voice signal based on a voice activity detection result corresponding to a first voice signal, and a noise reduction process is performed on the second voice signal based on the noise signal to obtain a voice signal enhancement result.

[0120] In one example, the preset first value may be a value greater than 1, for example, the preset first value may be 5. At this time, it may indicate that the noise in the collection environment where the voice signal to be processed is located is small, and does not affect the voice communication effect of the air conduction microphone.

[0121] In this case, the first processing mode can be characterized by: performing voice activity detection on the first voice signal collected by the bone conduction main microphone, so that when the voice activity detection result indicates that the bone conduction main microphone has not collected the voice signal of the current speaker, the second voice signal at this time can be determined as a noise signal. At this time, based on the noise signal, noise reduction processing can be performed on the second voice signal to obtain a voice signal enhancement result.

[0122] For example, Figure 4 The following is a flow chart of a speech enhancement method provided in an embodiment of the present application in an actual application scenario. Figure 4 As shown, the speech enhancement method provided in the present application can be applied to headphones, and the Bluetooth module of the headphones is connected to a terminal device (for example, Figure 4 The mobile phone shown in the figure) is used for voice communication.

[0123] like Figure 4 As shown, the processing flow corresponding to the first processing mode is: after the second voice signal is collected by the air conduction microphone and the first voice signal is collected by the bone conduction main microphone, voice activity detection can be performed on the first voice signal, and when it is determined based on the voice detection result that the bone conduction main microphone has not collected the voice signal of the current speaker, the noise signal in the second voice signal is determined.

[0124] Then, an adaptive filter is obtained to control the signal bandwidth information of the second speech signal during the noise reduction process through the adaptive filter, so as to perform noise reduction processing on the second speech signal based on the noise signal to obtain a speech signal enhancement result.

[0125] This implementation method can use the voice signal collected by the air conduction microphone as the voice signal to be processed in a low-noise environment, and perform voice activity detection through the voice signal collected by the bone conduction main microphone to estimate the noise signal, and perform noise reduction processing on the voice signal to be processed based on the obtained noise signal to achieve voice enhancement, which can avoid processing more voice signals, reduce the amount of signal calculation, and save power consumption.

[0126] Case 2: When it is determined that the first signal-to-noise ratio is less than a preset first value and greater than a preset second value, the second signal-to-noise ratio corresponding to the first speech signal is determined; when it is determined that the second signal-to-noise ratio is greater than a preset third value, the signal processing mode that matches the speech signal to be processed is determined to be the second processing mode.

[0127] The second processing mode represents a processing method for obtaining a speech signal enhancement result by determining a low-frequency speech signal based on the first speech signal and a high-frequency speech signal based on the second speech signal.

[0128] In one example, the preset second value may be a value less than 1, for example, the preset second value may be 0. At this time, it can be determined that the noise signal of the collection environment where the voice signal to be processed is located is large. At this time, although the air conduction microphone can collect the voice signal, due to the large noise in the collection environment, the voice signal collected by the air conduction microphone is easily distorted, which easily affects the voice communication effect. Therefore, the voice communication quality cannot be guaranteed based solely on the voice signal collected by the air conduction microphone.

[0129] In this case, the low-frequency voice signal can be determined based on the voice signal collected by the bone conduction main microphone (the bone conduction main microphone can collect low-frequency band voice signals and high-frequency band noise signals), and the high-frequency voice signal can be determined based on the voice signal collected by the air conduction microphone. After fusing the low-frequency voice signal and the high-frequency voice signal, a voice signal for communication with good communication quality (that is, a voice signal enhancement result) can be obtained.

[0130] In one example, the preset third value may be a value greater than 1, for example, the preset third value may be 10. At this time, it can be determined that the voice signal collected by the bone conduction main microphone has good quality. At this time, the voice signal collected by the bone conduction main microphone and the voice signal collected by the air conduction microphone may be fused to achieve voice enhancement.

[0131] Based on this, Figure 4 As shown, the speech enhancement processing flow corresponding to the second processing mode can be described as the following process.

[0132] First, based on a preset low-pass filter, the first speech signal and the third speech signal collected by the third microphone (i.e., the bone conduction reference microphone) are filtered, and based on the filtered third speech signal, the filtered first speech signal is denoised to obtain a low-frequency speech signal.

[0133] The preset low-pass filter is determined based on the frequency band distribution of the first speech signal. The third speech signal represents the noise of the collection environment where the speech signal to be processed is located. The third microphone (ie, the bone conduction reference microphone) can collect noise in the air.

[0134] Then, a high-pass filter is set based on the low-frequency voice signal. The high-pass filter is determined based on the signal bandwidth of the low-frequency voice signal. That is, the bandwidth of the high-pass filter can be determined based on the bandwidth of the low-frequency voice signal. For example, assuming that the bandwidth of the low-frequency voice signal is 0-2K Hz, then the bandwidth corresponding to the high-pass filter is greater than 2K Hz.

[0135] Next, based on a high-pass filter, the second voice signal and the third voice signal are filtered, and based on the filtered third voice signal, the filtered second voice signal is subjected to noise reduction processing to obtain a high-frequency voice signal.

[0136] In one example, after filtering the second voice signal through a high-pass filter, the third voice signal can also be filtered through the high-pass filter to obtain a high-frequency noise signal. Then, the filtered second voice signal can be subjected to noise reduction processing based on the high-frequency noise signal to obtain a high-frequency voice signal.

[0137] In one example, before performing noise reduction processing on the filtered second voice signal based on the high-frequency noise signal, an adaptive filter can be first obtained, and based on the adaptive filter, the high-frequency noise signal and the signal bandwidth information within the filtered second voice signal can be controlled to achieve noise reduction processing on the filtered second voice signal and obtain a high-frequency voice signal.

[0138] Finally, the signal bandwidth information of the low-frequency voice signal and the high-frequency voice signal can be controlled by an adaptive filter to accurately fuse the low-frequency voice signal and the high-frequency voice signal to obtain a fused voice signal. At this time, in order to align the residual noise in the low-frequency voice signal and the high-frequency voice signal, the fused voice signal can be subjected to noise reduction processing based on the third voice signal to obtain a voice signal enhancement result.

[0139] This implementation can determine the low-frequency voice signal based on the voice signal collected by the bone conduction reference microphone and the voice signal collected by the bone conduction main microphone in a high-noise environment, and determine the high-frequency voice signal based on the voice signal collected by the bone conduction reference microphone and the voice signal collected by the air conduction microphone, and then perform noise reduction processing on the voice signal after the fusion of the low-frequency voice signal and the high-frequency voice signal to obtain a voice signal enhancement result. In a high-noise environment, the bone conduction microphone can make up for the shortcomings of the air conduction microphone, and can effectively maintain the sound quality of the voice signal and improve the voice communication quality in a high-noise environment.

[0140] Case three: when it is determined that the first signal-to-noise ratio is less than a preset second value, or when it is determined that the second signal-to-noise ratio is less than a preset third value, the signal processing mode matching the speech signal to be processed is determined to be the third processing mode.

[0141] Among them, the third processing mode represents a processing method of performing noise reduction processing on the first voice signal based on the third voice signal collected by the third microphone to obtain a voice signal enhancement result.

[0142] In one example, when it is determined that the first signal-to-noise ratio is less than a preset second value (for example, the preset second value is 0), it means that the air conduction microphone is unavailable, or the air conduction microphone cannot collect the speech signal. In this case, a bone conduction microphone (for example, the above-mentioned bone conduction main microphone and bone conduction reference microphone) can be used to collect the speech signal to be processed, and speech enhancement processing can be performed on the speech signal to be processed.

[0143] In one example, when it is determined that the second signal-to-noise ratio is less than a preset third value, it means that the quality of the voice signal collected by the bone conduction main microphone is poor. At this time, the voice signal (i.e., the noise signal) collected by the bone conduction reference microphone can be used for noise reduction processing to obtain a voice signal enhancement result.

[0144] Based on this, Figure 4 As shown, the speech enhancement processing flow corresponding to the third processing mode is: based on the third speech signal (that is, the noise signal) collected by the bone conduction reference microphone, the first speech signal collected by the bone conduction main microphone is subjected to noise reduction processing to obtain a speech signal enhancement result. In this process, the signal bandwidth information of the first speech signal in the noise reduction process can be controlled by an adaptive filter to ensure the accuracy of the noise reduction process. After the noise reduction processing, a speech signal enhancement result is obtained.

[0145] This implementation can use the bone conduction microphone for voice communication in scenarios where the air conduction microphone cannot collect voice signals, or the air conduction microphone cannot be used / restored. This can effectively ensure the smoothness of voice communication when the air conduction microphone fails, solving the communication needs in scenarios where the air conduction microphone cannot collect voice signals.

[0146] S205: Perform noise reduction processing on the speech signal to be processed according to the matched signal processing mode to obtain a speech signal enhancement result.

[0147] Figure 5 The following is a schematic diagram of an application flow of a speech enhancement method provided in an embodiment of the present application. Figure 5 As shown in FIG. 1 , after acquiring the speech signal to be processed, the first speech signal acquired by the second main microphone (i.e., the bone conduction main microphone mentioned above) can be subjected to speech activity detection to determine the noise situation of the collection environment where the speech signal to be processed is located. At this time, scene recognition can be performed by calculating the first signal-to-noise ratio and the second signal-to-noise ratio (i.e., Figure 5 scene recognizer shown).

[0148] For example, when it is determined that the first signal-to-noise ratio is greater than a preset first value, the scene recognition result is determined to be scene 1 (at this time, scene 1 corresponds to the above-mentioned situation 1). At this time, the noise signal in the second voice signal collected by the first microphone (such as the above-mentioned air conduction microphone) can be determined by performing voice activity detection on the first voice signal collected by the second microphone (such as the above-mentioned bone conduction main microphone). At this time, the signal to be processed (that is, the second voice signal) can be subjected to noise reduction processing based on the noise signal to obtain a voice signal enhancement result.

[0149] For another example, when it is determined that the first signal-to-noise ratio is less than a preset first value, and the first signal-to-noise ratio is greater than a preset second value, and the second signal-to-noise ratio is greater than a preset third value, the scene recognition result is determined to be scene 2 (at this time, scene 2 corresponds to the above situation 2). At this time, the first voice signal can be filtered by a preset low-pass filter, and then, based on the third voice signal collected by the third microphone (such as the above-mentioned bone reference microphone), the filtered first voice signal is subjected to noise reduction processing to obtain a low-frequency voice signal. Then, a high-pass filter can be set based on the frequency band distribution of the low-frequency voice signal. At this time, the second voice signal and the third voice signal can be filtered based on the high-pass filter, and the filtered second voice signal can be subjected to noise reduction processing based on the filtered third voice signal to obtain a high-frequency voice signal. Then, the low-frequency voice signal and the high-frequency voice signal can be fused to obtain a fused voice signal. Finally, the fused voice signal can be subjected to noise reduction processing to obtain a voice signal enhancement result.

[0150] For another example, when it is determined that the first signal-to-noise ratio is less than a preset second value, or when it is determined that the second signal-to-noise ratio is less than a preset third value, the scene recognition result is determined to be scene 3 (in this case, scene 3 corresponds to the above-mentioned situation 3). At this time, the noise signal can be determined by the third voice signal, and the first voice signal can be subjected to noise reduction processing based on the noise signal to obtain a voice signal enhancement result.

[0151] Those skilled in the art will appreciate that, in the above method of specific implementation, the order in which the steps are written does not imply a strict execution order and does not constitute any limitation on the implementation process. The specific execution order of the steps should be determined by their functions and possible internal logic.

[0152] Figure 6 A schematic diagram of the structure of a speech enhancement device provided in an embodiment of the present application is shown in FIG. Figure 6 As shown, the speech enhancement device 600 includes:

[0153] The acquisition unit 601 is used to acquire a speech signal to be processed; wherein the speech signal to be processed represents speech signals collected by multiple microphones.

[0154] The determination unit 602 is used to determine a signal processing mode that matches the speech signal to be processed based on the speech signal to be processed; wherein the signal processing mode represents a processing method for enhancing the speech signal to be processed.

[0155] The processing unit 603 is used to perform noise reduction processing on the speech signal to be processed according to the matched signal processing mode to obtain a speech signal enhancement result.

[0156] Figure 7 A structural diagram of another speech enhancement device provided in an embodiment of the present application is shown in FIG. Figure 7 As shown, the speech enhancement device 700 includes:

[0157] The acquisition unit 701 is used to acquire a speech signal to be processed; wherein the speech signal to be processed represents speech signals collected by multiple microphones.

[0158] The determination unit 702 is used to determine a signal processing mode that matches the speech signal to be processed based on the speech signal to be processed; wherein the signal processing mode represents a processing method for enhancing the speech signal to be processed.

[0159] The processing unit 703 is used to perform noise reduction processing on the speech signal to be processed according to the matched signal processing mode to obtain a speech signal enhancement result.

[0160] In one example, the plurality of microphones include a first microphone for collecting voice signals transmitted through air, a second microphone for collecting voice signals transmitted through bones, and a third microphone for collecting noise in the air.

[0161] In one example, the determining unit 702 is configured to:

[0162] Based on the first voice signal collected by the second microphone, voice activity detection is performed to obtain a voice activity detection result; wherein the voice activity detection result indicates whether the second microphone collects the voice signal of the current speaker;

[0163] Based on the voice activity detection result, a noise signal is determined from the second voice signal collected by the first microphone, and a first signal-to-noise ratio of the second voice signal is determined; wherein the noise signal represents the noise of the collection environment where the signal to be processed is located;

[0164] Based on the first signal-to-noise ratio, a signal processing mode matching the speech signal to be processed is determined.

[0165] In one example, the determining unit 702 includes:

[0166] The first sub-determining module 7021 is used to determine that the signal processing mode matching the speech signal to be processed is the first processing mode when it is determined that the first signal-to-noise ratio is greater than a preset first value;

[0167] Among them, the first processing mode represents a processing method in which a noise signal is determined from a second voice signal based on a voice activity detection result corresponding to a first voice signal, and a noise reduction process is performed on the second voice signal based on the noise signal to obtain a voice signal enhancement result.

[0168] In one example, the determining unit 702 includes:

[0169] The second sub-determination module 7022 is used to determine a second signal-to-noise ratio corresponding to the first speech signal when it is determined that the first signal-to-noise ratio is less than a preset first value and greater than a preset second value;

[0170] When it is determined that the second signal-to-noise ratio is greater than a preset third value, determining that the signal processing mode matching the speech signal to be processed is the second processing mode;

[0171] The second processing mode represents a processing method for obtaining a speech signal enhancement result by determining a low-frequency speech signal based on the first speech signal and a high-frequency speech signal based on the second speech signal.

[0172] In one example, the second sub-determination module 7022 is used to:

[0173] Based on a preset low-pass filter, the first voice signal and the third voice signal collected by the third microphone are filtered, and based on the filtered third voice signal, the filtered first voice signal is subjected to noise reduction processing to obtain a low-frequency voice signal; wherein the third voice signal represents the noise of the collection environment where the voice signal to be processed is located;

[0174] Based on the low-frequency speech signal, a high-pass filter is set; wherein the high-pass filter is determined based on the signal bandwidth of the low-frequency speech signal;

[0175] Based on a high-pass filter, the second voice signal and the third voice signal are filtered, and based on the filtered third voice signal, the filtered second voice signal is denoised to obtain a high-frequency voice signal;

[0176] The low-frequency speech signal and the high-frequency speech signal are fused, and the fused speech signal is subjected to noise reduction processing to obtain the speech signal enhancement result.

[0177] In one example, the determining unit 702 includes:

[0178] The third sub-determination module 7023 is used to determine that the signal processing mode matching the speech signal to be processed is the third processing mode when it is determined that the first signal-to-noise ratio is less than a preset second value, or when it is determined that the second signal-to-noise ratio is less than a preset third value;

[0179] Among them, the third processing mode represents a processing method of performing noise reduction processing on the first voice signal based on the third voice signal collected by the third microphone to obtain a voice signal enhancement result.

[0180] In one example, the processing unit 703 is configured to:

[0181] Acquire an adaptive filter; wherein the adaptive filter is used to control signal bandwidth information of the speech signal to be processed during the noise reduction process;

[0182] Based on the adaptive filter and the matching signal processing mode, the speech signal to be processed is subjected to noise reduction processing to obtain the speech signal enhancement result.

[0183] Figure 8 A schematic diagram of the structure of a computer device provided in an embodiment of the present application is shown in FIG. Figure 8 As shown, the computer device 800 includes: a memory 801 and a processor 802 .

[0184] Memory 801 is a memory used to store instructions executable by processor 802 .

[0185] The processor 802 is configured to execute the method provided in the above embodiment.

[0186] The computer device further includes a receiver 803 and a transmitter 804. The receiver 803 is used to receive instructions and data sent by an external device, and the transmitter 804 is used to send instructions and data to the external device.

[0187] Fig. 9 It is a block diagram of a computer device according to an exemplary embodiment. The device may be a mobile phone, a computer, a digital broadcast terminal, a messaging device, a game console, a tablet device, a medical device, a fitness device, a personal digital assistant, etc.

[0188] Device 900 may include one or more of the following components: a processing component 902 , a memory 904 , a power component 906 , a multimedia component 908 , an audio component 910 , an input / output (I / O) interface 912 , a sensor component 914 , and a communication component 916 .

[0189] The processing component 902 generally controls the overall operation of the device 900, such as operations associated with display, phone calls, data communications, camera operations, and recording operations. The processing component 902 may include one or more processors 920 to execute instructions to complete all or part of the steps of the above-mentioned method. In addition, the processing component 902 may include one or more modules to facilitate the interaction between the processing component 902 and other components. For example, the processing component 902 may include a multimedia module to facilitate the interaction between the multimedia component 908 and the processing component 902.

[0190] The memory 904 is configured to store various types of data to support operations on the device 900. Examples of such data include instructions for any application or method operating on the device 900, contact data, phone book data, messages, pictures, videos, etc. The memory 904 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0191] The power supply component 906 provides power to the various components of the device 900. The power supply component 906 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power for the device 900.

[0192] The multimedia component 908 includes a screen that provides an output interface between the device 900 and the user. In some embodiments, the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive input signals from the user. The touch panel includes one or more touch sensors to sense touch, slide, and gestures on the touch panel. The touch sensor may not only sense the boundaries of the touch or slide action, but also detect the duration and pressure associated with the touch or slide operation. In some embodiments, the multimedia component 908 includes a front camera and / or a rear camera. When the device 900 is in an operating mode, such as a shooting mode or a video mode, the front camera and / or the rear camera may receive external multimedia data. Each front camera and rear camera may be a fixed optical lens system or have a focal length and optical zoom capability.

[0193] The audio component 910 is configured to output and / or input audio signals. For example, the audio component 910 includes a microphone (MIC), and when the device 900 is in an operating mode, such as a call mode, a recording mode, and a speech recognition mode, the microphone is configured to receive an external audio signal. The received audio signal can be further stored in the memory 904 or sent via the communication component 916. In some embodiments, the audio component 910 also includes a speaker for outputting audio signals.

[0194] I / O interface 912 provides an interface between processing component 902 and peripheral interface modules, such as keyboards, click wheels, buttons, etc. These buttons may include but are not limited to: a home button, a volume button, a start button, and a lock button.

[0195] The sensor assembly 914 includes one or more sensors for providing various aspects of status assessment for the device 900. For example, the sensor assembly 914 can detect the open / closed state of the device 900, the relative positioning of components, such as the display and keypad of the device 900, and the sensor assembly 914 can also detect the position change of the device 900 or a component of the device 900, the presence or absence of user contact with the device 900, the orientation or acceleration / deceleration of the device 900, and the temperature change of the device 900. The sensor assembly 914 may include a proximity sensor configured to detect the presence of nearby objects without any physical contact. The sensor assembly 914 may also include an optical sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, the sensor assembly 914 may also include an accelerometer, a gyroscope sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.

[0196] The communication component 916 is configured to facilitate wired or wireless communication between the device 900 and other devices. The device 900 can access a wireless network based on a communication standard, such as WiFi, 2G or 3G, or a combination thereof. In an exemplary embodiment, the communication component 916 receives a broadcast signal or broadcast-related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component 916 also includes a near field communication (NFC) module to facilitate short-range communication. For example, the NFC module can be implemented based on radio frequency identification (RFID) technology, infrared data association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology and other technologies.

[0197] In an exemplary embodiment, the apparatus 900 may be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the above method.

[0198] In an exemplary embodiment, a non-transitory computer-readable storage medium including instructions is also provided, such as a memory 904 including instructions, and the instructions can be executed by the processor 920 of the device 900 to perform the above method. For example, the non-transitory computer-readable storage medium can be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, etc.

[0199] A non-transitory computer-readable storage medium, when the instructions in the storage medium are executed by a processor of a terminal device, enables the terminal device to perform the above-mentioned speech enhancement method.

[0200] The embodiment of the present application also provides a computer program product, which carries computer execution instructions. The instructions included in the computer execution instructions can be used to execute the steps of the speech enhancement method in the above method embodiment. For details, please refer to the above method embodiment, which will not be repeated here.

[0201] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the invention disclosed herein. The present application is intended to cover any modification, use or adaptation of the present application, which follows the general principles of the present application and includes common knowledge or customary techniques in the art that are not disclosed in the present application. The specification and examples are intended to be exemplary only, and the true scope and spirit of the present application are indicated by the following claims.

[0202] It should be understood that the present application is not limited to the precise structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present application is limited only by the appended claims.

Claims

1. A speech enhancement method, characterized in that: include: Acquire a speech signal to be processed; wherein the speech signal to be processed represents speech signals collected by multiple microphones; Based on the speech signal to be processed, determining a signal processing mode that matches the speech signal to be processed; wherein the signal processing mode represents a processing method for enhancing the speech signal to be processed; According to the matched signal processing mode, the speech signal to be processed is subjected to noise reduction processing to obtain a speech signal enhancement result.

2. The method according to claim 1, characterized in that: The plurality of microphones include a first microphone for collecting voice signals transmitted through air, a second microphone for collecting voice signals transmitted through bones, and a third microphone for collecting noise in the air; The step of determining a signal processing mode that matches the speech signal to be processed based on the speech signal to be processed includes: Based on the first voice signal collected by the second microphone, voice activity detection is performed to obtain a voice activity detection result; wherein the voice activity detection result indicates whether the second microphone collects the voice signal of the current speaker; Based on the voice activity detection result, a noise signal is determined from the second voice signal collected by the first microphone, and a first signal-to-noise ratio of the second voice signal is determined; wherein the noise signal represents the noise of the collection environment where the signal to be processed is located; Based on the first signal-to-noise ratio, a signal processing mode matching the speech signal to be processed is determined.

3. The method according to claim 2, characterized in that The step of determining a signal processing mode matching the speech signal to be processed based on the first signal-to-noise ratio includes: When it is determined that the first signal-to-noise ratio is greater than a preset first value, determining that a signal processing mode matching the speech signal to be processed is a first processing mode; Among them, the first processing mode represents a processing method in which, based on the voice activity detection result corresponding to the first voice signal, a noise signal is determined from the second voice signal, and the second voice signal is subjected to noise reduction processing based on the noise signal to obtain a voice signal enhancement result.

4. The method according to claim 3, characterized in that The step of determining a signal processing mode matching the speech signal to be processed based on the first signal-to-noise ratio includes: When it is determined that the first signal-to-noise ratio is less than the preset first value and greater than a preset second value, determining a second signal-to-noise ratio corresponding to the first speech signal; In the case where it is determined that the second signal-to-noise ratio is greater than a preset third value, determining that the signal processing mode matching the speech signal to be processed is the second processing mode; The second processing mode represents a processing method for obtaining a speech signal enhancement result based on a low-frequency speech signal determined based on the first speech signal and a high-frequency speech signal determined based on the second speech signal.

5. The method according to claim 4, characterized in that The low-frequency speech signal determined based on the first speech signal and the high-frequency speech signal determined based on the second speech signal are used to obtain a speech signal enhancement result, including: Based on a preset low-pass filter, the first voice signal and the third voice signal collected by the third microphone are filtered, and based on the filtered third voice signal, the filtered first voice signal is subjected to noise reduction processing to obtain the low-frequency voice signal; wherein the third voice signal represents the noise of the collection environment where the voice signal to be processed is located; Based on the low-frequency voice signal, a high-pass filter is set; wherein the high-pass filter is determined based on the signal bandwidth of the low-frequency voice signal; Based on the high-pass filter, the second voice signal and the third voice signal are filtered, and based on the filtered third voice signal, the filtered second voice signal is denoised to obtain the high-frequency voice signal; The low-frequency speech signal and the high-frequency speech signal are fused, and the fused speech signal is subjected to noise reduction processing to obtain the speech signal enhancement result.

6. The method according to claim 4, characterized in that The step of determining a signal processing mode matching the speech signal to be processed based on the first signal-to-noise ratio includes: When it is determined that the first signal-to-noise ratio is less than the preset second value, or when it is determined that the second signal-to-noise ratio is less than the preset third value, determining that the signal processing mode matching the speech signal to be processed is the third processing mode; The third processing mode represents a processing method of performing noise reduction processing on the first voice signal based on the third voice signal collected by the third microphone to obtain a voice signal enhancement result.

7. The method according to any one of claims 1 to 6, characterized in that The step of performing noise reduction processing on the speech signal to be processed according to the matched signal processing mode to obtain a speech signal enhancement result includes: Acquire an adaptive filter; wherein the adaptive filter is used to control the signal bandwidth information of the speech signal to be processed during the noise reduction process; Based on the adaptive filter and the matching signal processing mode, the speech signal to be processed is subjected to noise reduction processing to obtain the speech signal enhancement result.

8. A speech enhancement device, characterized in that: include: An acquisition unit, configured to acquire a speech signal to be processed; wherein the speech signal to be processed represents speech signals collected by a plurality of microphones; A determination unit, configured to determine, based on the speech signal to be processed, a signal processing mode that matches the speech signal to be processed; wherein the signal processing mode represents a processing method for enhancing the speech signal to be processed; The processing unit is used to perform noise reduction processing on the speech signal to be processed according to the matched signal processing mode to obtain a speech signal enhancement result.

9. A computer device, characterized in that: include: A processor, and a memory communicatively connected to the processor; The memory stores computer-executable instructions; The processor executes the computer-executable instructions stored in the memory to implement the speech enhancement method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, which are used to implement the speech enhancement method according to any one of claims 1 to 7 when executed by a processor.

11. A computer program product, characterized in that The method comprises computer-executable instructions, which, when executed by a processor, implement the speech enhancement method according to any one of claims 1 to 7.

Images