Wake-up control method and apparatus, sound source position determination method and apparatus, and electronic device and vehicle

Abstract

A wake-up control method and apparatus, a sound source position determination method and apparatus, and an electronic device and a vehicle, which relate to the technical field of vehicle control. The wake-up control method comprises: acquiring voice information collected by a microphone device, and performing voice recognition on the voice information (101); if it is identified that the voice information includes a key word for waking up an intelligent device in a target space, parsing an audio signal corresponding to the voice information, so as to obtain sound source directivity information (102); on the basis of the sound source directivity information, determining whether the audio signal is propagated by means of multiple paths (103); and if the audio signal is propagated by means of the multiple paths, determining that a wake-up behavior corresponding to the voice information is a behavior of waking up from outside the target space the intelligent device in the target space, and performing controlling so as not to wake up the intelligent device in the target space (104). The method can accurately determine whether an intelligent device in a target space is waken up from outside the target space, so as to accurately perform corresponding suppression control; and there is no need to add a sound insulation material outside the target space, thereby saving on costs.

Description

Wake-up control method, sound source location determination method, device, electronic device, and vehicle

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application claims priority to the Chinese patent application filed with the China Patent Office on September 18, 2023, with application number 202311206626.3 and application name “Control method, device, electronic device and vehicle for in-vehicle voice assistant”, the entire contents of which are incorporated by reference into this application. Technical Field

[0003] The present application relates to the field of vehicle control technology, and is related to, but not limited to, a wake-up control method, a sound source location determination method, a device, an electronic device, and a vehicle. Background Art

[0004] With the rapid development of vehicle technology, in-car voice assistants receive user voice commands to control certain vehicle functions, helping users solve problems and providing significant convenience while driving. However, since in-car voice assistants are activated by voice commands, external noise can cause them to be falsely activated, necessitating a process to prevent them from being activated.

[0005] Currently, additional sound insulation can be added to the vehicle body to minimize external noise and reduce the likelihood of the voice assistant being activated. However, this approach not only increases costs, but also increases the vehicle's weight, reducing its range.

[0006] Summary of the Invention

[0007] In view of this, the present application provides a wake-up control method, a sound source location determination method, a device, an electronic device and a vehicle. The main purpose is to improve the current method of adding sound insulation materials to the vehicle body to isolate the sound from outside the vehicle from interfering with the smart devices in the vehicle. This will increase costs, and these sound insulation materials will increase the vehicle's own weight, resulting in a technical problem of reducing the vehicle's endurance.

[0008] In a first aspect, the present application provides a wake-up control method, comprising:

[0009] Acquire voice information collected by the microphone device and perform voice recognition on the voice information;

[0010] If it is recognized that the voice information contains a keyword for waking up the smart device in the target space, the audio signal corresponding to the voice information is analyzed to obtain sound source directionality information;

[0011] Based on the sound source directivity information, determining whether the audio signal is multipath propagated;

[0012] If the audio signal is multi-path propagated, it is determined that the wake-up behavior corresponding to the voice information is a behavior of waking up the smart device in the target space outside the target space, and the smart device in the target space is controlled not to be woken up.

[0013] Optionally, parsing the audio signal corresponding to the voice information to obtain sound source directionality information includes:

[0014] Calculating a data covariance matrix based on the audio signal;

[0015] De-noising the data covariance matrix to obtain an estimated signal covariance matrix;

[0016] Performing eigendecomposition and calculation on the estimated signal covariance matrix to obtain a main eigenvalue concentration, and using the main eigenvalue concentration as the sound source directivity information.

[0017] Optionally, the determining, based on the sound source directionality information, whether the audio signal is multipath propagated includes:

[0018] Comparing the main eigenvalue concentration with a preset threshold value;

[0019] If the main eigenvalue concentration is less than the preset threshold, it is determined that the audio signal is multipath propagated.

[0020] Optionally, calculating a data covariance matrix based on the audio signal includes:

[0021] Extracting signal data of a time period during which the keyword appears in the audio signal;

[0022] Calculate the signal data to obtain a data covariance matrix.

[0023] Optionally, performing noise removal processing on the data covariance matrix to obtain an estimated signal covariance matrix includes:

[0024] A noise covariance matrix corresponding to the estimated noise energy of the audio signal is subtracted from the data covariance matrix to obtain an estimated signal covariance matrix.

[0025] Optionally, performing eigendecomposition on the estimated signal covariance matrix and calculating to obtain a main eigenvalue concentration includes:

[0026] Calculating the audio signal to obtain signal-to-noise ratios at different frequencies;

[0027] Selecting a target frequency point whose signal-to-noise ratio is greater than a preset threshold among the different frequency points;

[0028] Performing eigendecomposition on the estimated signal covariance matrix corresponding to the target frequency point;

[0029] Based on the matrix eigenvalues ​​obtained by eigendecomposition, the main eigenvalue concentration is calculated.

[0030] Optionally, after determining whether the audio signal is multipath propagated based on the sound source directionality information, the method further includes:

[0031] If the audio signal is propagated in a single direct path, it is determined that the wake-up behavior corresponding to the voice information is a behavior of waking up the smart device in the target space, and control is performed to wake up the smart device in the target space.

[0032] Optionally, the target space includes a vehicle, and if it is recognized that the voice information contains a keyword for waking up a smart device in the target space, parsing an audio signal corresponding to the voice information to obtain sound source directionality information includes:

[0033] If it is recognized that the voice information contains a keyword for waking up the smart device in the car, the audio signal corresponding to the voice information is analyzed to obtain sound source directionality information;

[0034] Correspondingly, if the audio signal is multipath propagated, determining that the wake-up behavior corresponding to the voice information is a behavior of waking up the smart device in the target space from outside the target space, and controlling not to wake up the smart device in the target space, includes:

[0035] If the audio signal is multi-path propagated, it is determined that the wake-up behavior corresponding to the voice information is a behavior of waking up the in-vehicle smart device outside the vehicle, and the in-vehicle smart device is controlled not to be woken up.

[0036] Optionally, the target space includes an indoor space. If it is recognized that the voice information contains a keyword for waking up a smart device in the target space, parsing an audio signal corresponding to the voice information to obtain sound source directionality information includes:

[0037] If it is recognized that the voice information contains a keyword for waking up the indoor smart device, the audio signal corresponding to the voice information is analyzed to obtain sound source directionality information;

[0038] Correspondingly, if the audio signal is multipath propagated, determining that the wake-up behavior corresponding to the voice information is a behavior of waking up the smart device in the target space from outside the target space, and controlling not to wake up the smart device in the target space, includes:

[0039] If the audio signal is multi-path propagated, it is determined that the wake-up behavior corresponding to the voice information is an outdoor wake-up behavior of the indoor smart device, and the indoor smart device is controlled not to be woken up.

[0040] In a second aspect, the present application provides a method for determining a sound source location, comprising:

[0041] Acquire voice information collected by the microphone device and perform voice recognition on the voice information;

[0042] Analyzing the audio signal corresponding to the voice information to obtain sound source directionality information;

[0043] Based on the sound source directivity information, a sound source position of the audio signal is determined.

[0044] In a third aspect, the present application provides a device for determining a sound source position, comprising:

[0045] The first acquisition part is configured to acquire voice information collected by the microphone device and perform voice recognition on the voice information;

[0046] A first parsing part is configured to parse the audio signal corresponding to the voice information to obtain sound source directionality information;

[0047] The determining part is configured to determine the sound source position of the audio signal based on the sound source directivity information.

[0048] In a fourth aspect, the present application provides a wake-up control device, comprising:

[0049] The second acquisition part is configured to acquire voice information collected by the microphone device and perform voice recognition on the voice information;

[0050] The second parsing part is configured to parse the audio signal corresponding to the voice information to obtain sound source directionality information if it is recognized that the voice information contains a keyword for waking up the smart device in the target space;

[0051] a judging part, configured to judge whether the audio signal is multipath propagated based on the sound source directivity information;

[0052] The control part is configured to determine that the wake-up behavior corresponding to the voice information is a behavior of waking up the smart device in the target space outside the target space if the audio signal is multi-path propagated, and control not to wake up the smart device in the target space.

[0053] Optionally, the second analysis part is specifically configured to calculate a data covariance matrix based on the audio signal; perform noise removal processing on the data covariance matrix to obtain an estimated signal covariance matrix; perform eigendecomposition and calculation on the estimated signal covariance matrix to obtain a main eigenvalue concentration, and use the main eigenvalue concentration as the sound source directionality information.

[0054] Optionally, the judgment part is specifically configured to compare the main eigenvalue concentration with a preset threshold value; if the main eigenvalue concentration is less than the preset threshold value, it is determined that the audio signal is multipath propagated.

[0055] Optionally, the second parsing part is further configured to extract signal data of the time period in which the keyword appears in the audio signal; calculate the signal data to obtain a data covariance matrix; and convert the data covariance matrix to obtain a noisy data covariance matrix.

[0056] Optionally, the second analyzing part is further configured to subtract a noise covariance matrix corresponding to the estimated noise energy of the audio signal from the data covariance matrix to obtain an estimated signal covariance matrix.

[0057] Optionally, the second analysis part is further configured to calculate the audio signal to obtain the signal-to-noise ratio of different frequency points; select a target frequency point whose signal-to-noise ratio is greater than a preset threshold among the different frequency points; perform eigendecomposition on the signal covariance matrix estimated corresponding to the target frequency point; and calculate the main eigenvalue concentration based on the matrix eigenvalue obtained by the eigendecomposition.

[0058] Optionally, the judgment part is further configured to determine that the wake-up behavior corresponding to the voice information is an act of waking up the smart device in the target space if the audio signal is propagated in a single direct path, and control the waking up of the smart device in the target space.

[0059] Optionally, the target space includes the interior of a vehicle.

[0060] The second parsing part is specifically configured to parse the audio signal corresponding to the voice information to obtain sound source directionality information if it is recognized that the voice information contains a keyword for waking up the smart device in the car;

[0061] Correspondingly, the control part is specifically configured to determine that the wake-up behavior corresponding to the voice information is the behavior of waking up the smart device in the car from outside the car if the audio signal is multi-path propagated, and control not to wake up the smart device in the car.

[0062] Optionally, the target space includes indoors.

[0063] The second parsing part is specifically configured to parse the audio signal corresponding to the voice information to obtain sound source directionality information if it is recognized that the voice information contains a keyword for waking up the indoor smart device;

[0064] Correspondingly, the control part is specifically configured to determine that the wake-up behavior corresponding to the voice information is an outdoor wake-up behavior of the indoor smart device if the audio signal is multi-path propagated, and control not to wake up the indoor smart device.

[0065] In a fifth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a first processor, the method described in the first aspect is implemented. When the computer program is executed by a second processor, the method described in the second aspect is implemented.

[0066] In a sixth aspect, the present application provides an electronic device, comprising a storage medium, a first processor, a second processor, a computer program stored on the storage medium and runnable on the first processor, and a computer program stored on the storage medium and runnable on the second processor, wherein the first processor implements the method described in the first aspect when executing the computer program, and the second processor implements the method described in the second aspect when executing the computer program.

[0067] In a seventh aspect, the present application provides a vehicle comprising: the apparatus as described in the third and fourth aspects or the electronic device as described in the sixth aspect.

[0068] By means of the above technical solution, a wake-up control method, a sound source location determination method, a device, an electronic device, and a vehicle are provided. Compared with the current existing technology, the present application can determine whether the wake-up of the smart device in the target space comes from outside the target space only through the microphone device in the target space when the doors and windows of the target space are closed, and suppress it if so. Specifically, first, the voice information collected by the microphone device is obtained, and the voice information is subjected to voice recognition; if the voice information is recognized to contain keywords for waking up the smart device in the target space, the audio signal corresponding to the voice information is parsed to obtain the sound source directionality information; then, based on the sound source directionality information, it is determined whether the audio signal is multi-path propagated; if the audio signal is multi-path propagated, it is determined that the wake-up behavior corresponding to the voice information is the behavior of waking up the smart device in the target space from outside the target space, and the smart device in the target space is controlled not to be awakened. By applying the technical solution of the present application, it is possible to accurately determine whether the wake-up of the smart device in the target space comes from outside the target space, and then accurately perform corresponding suppression control. In addition, this technical solution does not require the addition of sound insulation materials outside the target space, which can save costs. When the target space is specifically the space inside the vehicle, it will not increase the vehicle's own weight, thereby avoiding affecting the vehicle's endurance.

[0069] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] 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.

[0071] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0072] FIG1 is a schematic diagram showing a flowchart of a wake-up control method provided in an embodiment of the present application;

[0073] FIG2 is a schematic diagram showing an application scenario provided by an embodiment of the present application;

[0074] FIG3 shows a schematic diagram of another application scenario provided by an embodiment of the present application;

[0075] FIG4 shows a flow chart of another wake-up control method provided in an embodiment of the present application;

[0076] FIG5 is a schematic diagram showing a process according to an embodiment of the present application;

[0077] FIG6 is a schematic diagram showing a flow chart of a method for determining a sound source position provided in an embodiment of the present application;

[0078] FIG7 shows a schematic structural diagram of a device for determining a sound source position according to an embodiment of the present application;

[0079] FIG8 shows a schematic structural diagram of a wake-up control device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0080] In order to be able to more clearly understand the above-mentioned objects, features and advantages of the present application, the scheme of the present application will be further described below. Although certain embodiments of the present application are shown in the accompanying drawings, it should be understood that the present application can be implemented in various forms and should not be construed as being limited to the embodiments set forth herein. On the contrary, these embodiments are provided to provide a more thorough and complete understanding of the present application. It should be understood that the drawings and embodiments of the present application are only for illustrative purposes and are not intended to limit the scope of protection of the present application. All other embodiments obtained by those of ordinary skill in the art without making creative work are within the scope of protection of the present application. It should be noted that, in the absence of conflict, the embodiments of the present application and the features in the embodiments can be combined with each other.

[0081] This embodiment provides a wake-up control method, as shown in FIG1 , including:

[0082] Step 101: Acquire voice information collected by a microphone device and perform voice recognition on the voice information.

[0083] The microphone device may include a distributed microphone array, where the microphones are dispersedly arranged in the space where the sound source is located, and the distance between the sound source and the microphones is at a similar order of magnitude as the distance between the microphones. It should be noted that this embodiment has no limitation on the arrangement of the microphone device.

[0084] Microphone devices have the function of collecting voice information; voice information may include voice information within the target space and voice information outside the target space, such as voice information inside the car and voice information outside the car, and voice information indoors and outdoors; voice recognition is a high-tech technology that allows machines to convert voice signals into corresponding text or commands through recognition and understanding. It is widely used in various fields such as industry, home appliances, communications, automotive electronics, medical care, home services, consumer electronics, etc.

[0085] In a specific application scenario, when obtaining voice information collected by a microphone device and performing voice recognition on the voice information, the method of this embodiment may include: using a communication device to obtain the voice information collected by the microphone device, and using a voice recognition tool to perform voice recognition on the voice information. This embodiment does not limit this.

[0086] Step 102: If it is recognized that the voice information contains a keyword for waking up the smart device in the target space, the audio signal corresponding to the voice information is analyzed to obtain sound source directionality information.

[0087] In an embodiment of the present application, if the target space includes the interior of a vehicle, and if the voice information is recognized to contain keywords for waking up a smart device (such as a voice assistant) in the target space, then the audio signal corresponding to the voice information is parsed to obtain the sound source directionality information, including: if the voice information is recognized to contain keywords for waking up a voice assistant in the vehicle, then the audio signal corresponding to the voice information is parsed to obtain the sound source directionality information. If the target space includes an indoor space, and if the voice information is recognized to contain keywords for waking up a smart device (such as a voice assistant) in the target space, then the audio signal corresponding to the voice information is parsed to obtain the sound source directionality information, including: if the voice information is recognized to contain keywords for waking up a voice assistant in the indoor space, then the audio signal corresponding to the voice information is parsed to obtain the sound source directionality information.

[0088] The target space includes indoors, in a vehicle, etc. Specifically, the target space may be a cabin space in a vehicle. The voice assistant may be a voice assistant in the vehicle or a voice assistant in the room.

[0089] The in-vehicle smart device (such as a voice assistant) may refer to an intelligent voice system that is combined with a voice assistant service to execute simple commands and significantly improve driving safety in order to avoid being distracted by reading screen content for a long time while driving. The keyword may be a pre-set word that triggers the in-vehicle smart device (such as a voice assistant) to start working. The keyword may specifically include a single word or multiple words. In a specific application scenario, if a voice message is recognized to contain a keyword that wakes up the in-vehicle smart device (such as a voice assistant), the audio signal corresponding to the voice message is parsed to obtain sound source directionality information. The method of this embodiment may include: if a voice message is recognized by the smart device (such as a voice assistant) or the vehicle host to contain a keyword that wakes up the in-vehicle smart device (such as a voice assistant), the audio signal corresponding to the voice message is parsed to obtain sound source directionality information. This embodiment is not limited to this.

[0090] Smart devices (such as voice assistants) can also be indoor smart devices (such as voice assistants). Indoor smart devices (such as voice assistants) can refer to smart voice systems that are combined with voice assistant services to facilitate the control of other indoor devices and can execute simple instructions, greatly improving the convenience of device control; keywords can be pre-set words that trigger indoor smart devices (such as voice assistants) to start working, and the keywords can specifically include a single word or multiple words. In a specific application scenario, if it is recognized that the voice information contains keywords that wake up the indoor smart device (such as a voice assistant), the audio signal corresponding to the voice information is parsed to obtain the sound source directionality information. The method of this embodiment may include: if the smart device (such as a voice assistant) or the server recognizes that the voice information contains keywords that wake up the indoor smart device (such as a voice assistant), the audio signal corresponding to the voice information is parsed to obtain the sound source directionality information. This embodiment does not limit this.

[0091] The directivity of a sound source can be measured as the transducer response in a specified plane passing through the center of the sound as a function of the direction of the emitted or incident sound wave, at a fixed frequency. When measuring the sound pressure level radiated by a sound source at a fixed distance from the source within a given frequency band, it is often found that the sound pressure level varies depending on the direction of the source. Low-frequency radiation from many noise sources is nearly non-directional, but its directivity increases with increasing frequency. This is because sound waves radiated from different parts of the vibration source arrive at different points in space at different times, resulting in interference and uneven directional radiation.

[0092] As shown in Figures 2 and 3, for wake-up inside the vehicle, due to the low reflection inside the vehicle, the microphone device mainly receives the audio signal of a single direct path. However, for wake-up outside the vehicle, due to the obstruction of the vehicle body and windows, no direct path can be formed. The sound can only be transmitted into the vehicle through the vibration of solid objects, thus transforming the point sound source into a surface sound source, resulting in a multipath propagation effect. Based on the two different propagation situations, this embodiment utilizes the characteristics of voice signal propagation to circumvent the disadvantage that the distributed array in the microphone device cannot directly utilize the high-frequency directional characteristics. This allows the use of distributed arrays alone to perform window wake-up judgment. That is, based on the audio signal analysis, the directionality of the sound source is determined based on the sound frequency and direction. The sound directionality is used to determine whether the audio signal is multipath propagation, and then distinguish whether the voice information comes from inside or outside the vehicle. This is used to determine whether the voice information is a user instruction inside the vehicle. Based on this analysis, smart devices (such as voice assistants) can determine whether to execute the corresponding operation of the voice information.

[0093] For example, for indoor wake-up, due to the small indoor reflection, the indoor microphone device mainly receives the audio signal of a single direct path; for outdoor wake-up, due to the obstruction of walls, doors and windows, no direct path can be formed, and the sound can only be transmitted into the room through the vibration of solid objects, thereby changing the point sound source into a surface sound source, resulting in a multi-path propagation effect. Based on the two different propagation situations, this embodiment utilizes the characteristics of voice signal propagation to avoid the disadvantage that the distributed array in the microphone device cannot directly utilize the high-frequency directional characteristics, so that the window wake-up judgment can be performed only by relying on the distributed array. That is, based on the audio signal, the directionality of the sound source is determined according to the sound frequency and direction. The sound directionality is used to determine whether the audio signal is multi-path propagation, and then distinguish whether the voice information comes from indoors or outdoors. It is used to determine whether the voice information is an instruction from the user indoors. Smart devices (such as voice assistants) can determine whether to execute the operation corresponding to the voice information based on the above analysis.

[0094] Step 103: Based on the sound source directivity information, determine whether the audio signal is multipath propagated.

[0095] In some examples, the audio signal may include propagation through a sound propagation medium within the target space (such as inside a car or indoors) and propagation through a sound propagation medium outside the target space. The sound propagation medium within the target space may be air propagation, and the sound propagation medium outside the target space may be window glass of the target space or a wall of the target space (or a structural shell constituting the target space) and air propagation, etc.; multi-path propagation may be sound propagation through multiple paths formed due to wake-up outside the target space. Among them, when the target space is inside a car, wake-up outside the car may be when a person outside the car wakes up an in-car smart device (such as a voice assistant) when the vehicle is stopped, the windows and doors are closed, and the in-car system is powered on. This embodiment does not limit this.

[0096] For an embodiment of the present application, when judging whether an audio signal is multi-path propagated based on the sound source directionality information, the method of this embodiment may include: judging whether the audio signal is multi-path propagated based on the sound source directionality information through an intelligent device (such as a voice assistant) or a vehicle host or a server, etc., and this embodiment does not limit this.

[0097] Step 104: If the audio signal is multi-path propagated, determine that the wake-up behavior corresponding to the voice information is a behavior of waking up the smart device in the target space from outside the target space, and control the smart device in the target space not to wake up.

[0098] For the embodiment of the present application, if the above-mentioned identification and analysis determine that the audio signal is multi-path propagation, then it can be determined that the current wake-up behavior is the behavior of waking up the smart device (such as a voice assistant) in the target space outside the target space. For example, at this time, the smart device (such as a voice assistant) in the target space can be controlled not to wake up.

[0099] For example, if the above-mentioned identification and analysis determines that the audio signal is multi-path propagation, then it can be determined that the current wake-up behavior is the behavior of waking up the smart device (such as a voice assistant) inside the car (or indoors) from outside the car (or outdoors). For example, at this time, the smart device (such as a voice assistant) inside the car (or indoors) can be controlled not to wake up.

[0100] In an embodiment of the present application, if the target space includes the interior of a vehicle and the audio signal is propagated through multiple paths, determining that the wake-up behavior corresponding to the voice information is an action of waking up a smart device in the target space from outside the target space, and controlling not to wake up the smart device in the target space, includes: if the audio signal is propagated through multiple paths, determining that the wake-up behavior corresponding to the voice information is an action of waking up a smart device in the vehicle from outside the vehicle, and controlling not to wake up the smart device in the vehicle. If the target space includes an indoor space and the audio signal is propagated through multiple paths, determining that the wake-up behavior corresponding to the voice information is an action of waking up a smart device in the target space from outside the target space, and controlling not to wake up the smart device in the target space, includes: if the audio signal is propagated through multiple paths, determining that the wake-up behavior corresponding to the voice information is an action of waking up a smart device in the indoor space from outside the vehicle, and controlling not to wake up the smart device in the indoor space.

[0101] The embodiment of the present application provides a wake-up control method, which can save costs and will not increase the weight of the overall structure constituting the target space. Specifically, first, the voice information collected by the microphone device is obtained, and the voice information is subjected to voice recognition; if it is recognized that the voice information contains keywords for waking up the smart device in the target space, the audio signal corresponding to the voice information is parsed to obtain the sound source directionality information; then, based on the sound source directionality information, it is determined whether the audio signal is multi-path propagated; if the audio signal is multi-path propagated, it is determined that the wake-up behavior corresponding to the voice information is the behavior of waking up the smart device in the target space from outside the target space, and the smart device in the target space is controlled not to be awakened. By applying the technical solution of the embodiment of the present application, it is possible to accurately determine whether the wake-up of the smart device in the target space comes from outside the target space, and then accurately perform corresponding suppression control, and this technical solution does not require the addition of sound insulation materials outside the target space, which can save costs and will not increase the weight of the overall structure constituting the target space.

[0102] The embodiment of the present application provides a wake-up control method. When the target space is specifically the interior of a vehicle, the vehicle can be stopped, the doors and windows are closed, and the in-vehicle system is powered on. The method can determine whether the wake-up of the in-vehicle smart device originates from outside the vehicle, and if so, suppress the wake-up. Specifically, the method first obtains the voice information collected by the microphone device and performs voice recognition on the voice information. If the voice information is found to contain keywords that wake up the in-vehicle smart device, the audio signal corresponding to the voice information is parsed to obtain sound source directionality information. Based on the sound source directionality information, the method then determines whether the audio signal is multipath propagated. If the audio signal is multipath propagated, the method determines that the wake-up behavior corresponding to the voice information is an outside wake-up behavior of the in-vehicle smart device, and controls the in-vehicle smart device not to wake up. By applying the technical solution of the present application, it is possible to accurately determine whether the wake-up of the in-vehicle smart device originates from outside the vehicle, and then accurately perform corresponding suppression control. In addition, the method does not require the addition of additional sound insulation materials to the vehicle body, which can save costs, and does not increase the vehicle's own weight, thereby avoiding affecting the vehicle's endurance.

[0103] Furthermore, in order to facilitate understanding of the implementation process of this embodiment, the principles involved in this embodiment are specifically described below. It should be noted that the following example content is only one optional method of the method of this embodiment, and this embodiment does not limit the specific implementation method of the method.

[0104] Regarding the wake-up methods of smart devices in the target space, they can be divided into wake-up within the target space and wake-up outside the target space.

[0105] For wake-up in the target space, since the reflection in the target space is small, the microphone mainly receives the audio signal of the direct path. The data model can be expressed as follows: x(t) = d θ *s(t)+n(t) (Formula 1)

[0106] In formula 1, x(t) represents the microphone received signal vector, d θ The steering vector represents the sound source position, s(t) represents the sound source signal, and n(t) represents the noise signal vector.

[0107] Furthermore, the data covariance matrix obtained for the wake-up data in the target space can be expressed as the following formula 2:

[0108] In Formula 2, R represents the covariance matrix in the single-path propagation scenario, that is, the data covariance matrix obtained by the wake-up data in the target space, E represents the mean operation, x(t) represents the microphone received signal vector, and d θ The steering vector representing the sound source position, σ 2represents the noise power, H represents the conjugate transpose operation, and I represents the identity matrix.

[0109] In specific application scenarios, when waking up outside the target space, due to the obstruction of the overall structure of the target space (in the case of a car, the overall structure of the target space can be the car body) and the windows of the target space, no direct path can be formed. The sound can only be transmitted to the target space through the vibration of the solid, thus changing the point sound source into a surface sound source, resulting in a multipath propagation effect. Assuming that there are P propagation paths, the data model can be expressed as follows:

[0110] In formula 3, xc(t) represents the microphone received signal vector in the multipath propagation scenario, that is, the data covariance matrix obtained by the wake-up data outside the target space, d θ The steering vector represents the sound source position, s(t) represents the sound source signal, and n(t) represents the noise signal vector.

[0111] Furthermore, the data covariance matrix obtained for the wake-up data outside the target space can be expressed as the following formula 4:

[0112] In formula 4, R c represents the covariance matrix in the multipath propagation scenario, E represents the mean operation, x c (t) represents the microphone received signal vector in the multipath propagation scenario, The steering vector representing the sound source position, σ 2 represents the noise power, H represents the conjugate transpose operation, and I represents the identity matrix.

[0113] In the above formula 4 and formula 5, there may be noise data. In this embodiment, the noise data may be removed to obtain an estimated signal covariance matrix.

[0114] For example, the estimated signal covariance matrix of the wakeup in the target space can be expressed as follows:

[0115] In formula 5, R s represents the estimated signal covariance matrix of arousal in the target space, R represents the covariance matrix, σ 2 represents the noise power, I represents the unit matrix, d θ represents the steering vector of the sound source position, and H represents the conjugate transpose operation.

[0116] The estimated signal covariance matrix of the wakeup outside the target space can be expressed as follows:

[0117] In formula six, The signal covariance matrix representing the estimate of arousal outside the target space, R c represents the covariance matrix, σ 2 represents the noise power, I represents the identity matrix, represents the steering vector of the sound source position, and H represents the conjugate transpose operation.

[0118] Based on Formulas 5 and 6, it can be analyzed that since the awakening in the target space is a direct path propagation, the obtained signal covariance matrix is ​​a rank-one matrix, while the signal covariance matrix rank of the awakening outside the target space is greater than 1. This feature can be used to distinguish between awakening in the target space and awakening outside the target space.

[0119] According to the above exemplary principles, this embodiment provides a specific method as shown in FIG4 , which includes the following steps:

[0120] Step 201: Acquire voice information collected by a microphone device and perform voice recognition on the voice information.

[0121] For the specific implementation process of the embodiment of the present application, please refer to the relevant description in step 101 of the embodiment, which will not be repeated here.

[0122] Step 202: Calculate and obtain a data covariance matrix based on the audio signal.

[0123] In the embodiment of the present application, the data covariance matrix may be a covariance matrix obtained from audio signal data.

[0124] Optionally, step 202 of the embodiment may specifically include: extracting signal data for the time period in which the keyword appears in the audio signal; and calculating the signal data to obtain a data covariance matrix. In this embodiment, regardless of whether the user is awakened within the target space or awakened outside the target space, a data covariance matrix is ​​calculated based on the audio information of the voice information, such as in the form of Formula 2 or Formula 4 above. Whether the user is awakened within the target space or awakened outside the target space can be determined based on further analysis results, such as by executing the process shown in steps 203 to 205.

[0125] Step 203: Perform noise removal on the data covariance matrix to obtain an estimated signal covariance matrix.

[0126] Optionally, step 203 may specifically include: subtracting a noise covariance matrix corresponding to the noise energy estimated for the audio signal from the data covariance matrix to obtain an estimated signal covariance matrix.

[0127] This embodiment can estimate noise energy and thereby obtain noise data in the audio signal, such as a noise covariance matrix corresponding to the estimated noise energy of the audio signal. The data covariance matrix is ​​then subjected to a noise energy removal process using the estimated noise energy. For example, the noise covariance matrix corresponding to the estimated noise energy of the audio signal is subtracted from the data covariance matrix to obtain an estimated signal covariance matrix. The specific form can be as shown in Formula 5 or Formula 6 above.

[0128] Step 204 : Perform eigendecomposition and calculation on the estimated signal covariance matrix to obtain the main eigenvalue concentration, and use the main eigenvalue concentration as the sound source directivity information.

[0129] Optionally, step 204 may specifically include: first calculating the audio signal to obtain the signal-to-noise ratio of different frequency points; then selecting a target frequency point whose signal-to-noise ratio is greater than a preset threshold among the different frequency points; then performing eigendecomposition on the estimated signal covariance matrix corresponding to the target frequency point, and then calculating the main eigenvalue concentration based on the matrix eigenvalues ​​obtained by the eigendecomposition.

[0130] In specific application scenarios, noise estimation can have certain deviations. If a frequency point with a relatively low signal-to-noise ratio is selected to obtain the corresponding estimated signal covariance matrix, when the matrix is ​​subjected to eigendecomposition, the corresponding principal eigenvalue is not stable, affecting the accuracy of subsequent evaluation. Therefore, this embodiment can select a frequency point with a signal-to-noise ratio greater than a certain threshold to obtain the corresponding estimated signal covariance matrix. When the matrix is ​​subjected to eigendecomposition, the corresponding principal eigenvalue is relatively stable. In addition, based on the concentration of this principal eigenvalue, accurate judgment of in-target and out-target space wake-up can be made.

[0131] When the estimated signal covariance matrix is ​​subjected to eigendecomposition, multiple eigenvalues ​​λ may be obtained, and the largest eigenvalue among them is selected as the main eigenvalue.

[0132] For example, the main eigenvalue concentration is obtained by using the following formula 7:

[0133] In Formula 7, ε is the main eigenvalue concentration, λ1 is the main eigenvalue, M is the number of matrix eigenvalues, and λi is the i-th eigenvalue.

[0134] Step 205: Compare the main eigenvalue concentration with a preset threshold value. If the main eigenvalue concentration is less than the preset threshold value, it is determined that the audio signal is multipath propagated.

[0135] In the embodiment of the present application, the absolute value obtained by subtracting the main eigenvalue concentration from 1 is compared with a preset threshold value to determine a reference value, and then the reference value is used to determine whether the audio signal is single-path propagation or multi-path propagation.

[0136] Optionally, through step 204 of the above embodiment, it can be obtained that in the case of awakening within the target space, the main eigenvalue concentration is close to 1, while in the case of awakening outside the target space, it is significantly less than 1, so that it can be distinguished whether it is awakening outside the target space. A threshold value is preset, the main eigenvalue concentration is subtracted from 1, and an absolute value analysis is performed to obtain a concentration difference. If the concentration difference is less than the preset threshold value, the voice wake-up signal is awakening within the target space; if the concentration difference is greater than the preset threshold value, the voice wake-up signal is awakening outside the target space. Among them, the threshold value is based on the difference between the wake-up within the target space and 1, and then absolute value processing is performed; and based on the difference between the wake-up outside the target space and 1, and then absolute value processing is performed, and then the absolute value corresponding to the above-mentioned wake-up within the target space and the absolute value corresponding to the wake-up outside the target space are comprehensively statistically analyzed to determine a threshold value.

[0137] In the embodiment of the present application, in the case of in-vehicle wake-up, the main eigenvalue concentration is close to 1, while in the case of outdoor wake-up, it is significantly less than 1, thus distinguishing whether it is an outdoor wake-up. A threshold value is preset, the main eigenvalue concentration is subtracted from 1, and an absolute value analysis is performed to obtain the concentration difference. If the concentration difference is less than the preset threshold value, the voice wake-up signal is an in-vehicle wake-up; if the concentration difference is greater than the preset threshold value, the voice wake-up signal is an outdoor wake-up. The threshold value is determined by subtracting the difference between the in-vehicle wake-up and 1, then performing absolute value processing; and subtracting the difference between the outdoor wake-up and 1, then performing absolute value processing, and then comprehensively calculating the absolute values ​​corresponding to the in-vehicle wake-up and the absolute values ​​corresponding to the outdoor wake-up. In the case of indoor wake-up, the main eigenvalue concentration is close to 1, while in the case of outdoor wake-up, it is significantly less than 1, thus distinguishing whether it is an outdoor wake-up. A threshold value is preset, the main eigenvalue concentration is subtracted from 1, and an absolute value analysis is performed to obtain the concentration difference. If the concentration difference is less than a preset threshold, the voice wake-up signal is an indoor wake-up; if the concentration difference is greater than a preset threshold, the voice wake-up signal is an outdoor wake-up. The threshold is determined by taking the absolute value of the difference between the indoor wake-up and 1, and then processing the absolute value; and taking the absolute value of the difference between the outdoor wake-up and 1, and then processing the absolute value. The absolute values ​​corresponding to the indoor and outdoor wake-ups are then combined to form a threshold.

[0138] For example, statistics are collected for the absolute value A corresponding to the subtraction of 1 from the in-vehicle (or indoor) wakeup and the absolute value B corresponding to the subtraction of 1 from the out-vehicle (or outdoor) wakeup. Absolute values ​​A include 0.3, 0.5, 0.2, 0.4, and 0.3, and absolute values ​​B include 0.6, 0.7, 0.5, 0.8, and 0.9. The average values ​​of absolute values ​​A and B are 0.34 and 0.7, respectively. Furthermore, as shown in the above example, the concentration of the main eigenvalues ​​for in-vehicle (or indoor) wakeup is close to 1, while the concentration of the main eigenvalues ​​for out-vehicle (or outdoor) wakeup is significantly less than 1. The average value of absolute values ​​A, 0.34, is less than the average value of absolute values ​​B, 0.7, indicating that these values ​​are valid. The quotient of 0.52, obtained by adding the average value of absolute values ​​A, 0.34, and the average value of absolute values ​​B, 0.7, and dividing by 2, is used as the threshold value. The current main eigenvalue concentration TZ is extracted as 0.45 and subtracted from 1. The absolute value of the difference obtained is 0.55, which is the concentration difference. The concentration difference 0.55 is greater than the threshold value 0.52, which determines that the audio signal is multipath propagation.

[0139] Through the specific implementation of the above embodiment, it is possible to determine whether the voice information is propagating through a single path or multiple paths. The single path propagation corresponds to wake-up within the target space, while the multipath propagation corresponds to wake-up outside the target space. Furthermore, the voice information can be determined to be within the target space or outside the target space, thereby performing relative processing on the voice information, thereby avoiding the impact on user experience and safety risks caused by incorrect operations. At the same time, wake-up failures caused by people in the target space waking up the smart device with low voices will not be caused.

[0140] Step 206: If the audio signal is multi-path propagated, determine that the wake-up behavior corresponding to the voice information is a behavior of waking up the smart device in the target space from outside the target space, and control the smart device in the target space not to wake up.

[0141] For the embodiment of the present application, in some application scenarios, it may also include: if the wake-up behavior is an behavior of waking up a smart device in the target space, then controlling the smart device in the target space to wake up.

[0142] Specifically, as shown in the schematic diagram of an application scenario in Figures 2 and 3, for wake-up inside the car, due to the small reflection inside the car, the microphone mainly receives the audio signal of the direct path; for wake-up outside the car, due to the obstruction of the car body and windows, no direct path can be formed, and the sound can only be transmitted into the car through the vibration of solids, thus changing the point sound source into a surface sound source, resulting in a multipath propagation effect.

[0143] The embodiments of the present application provide a wake-up control method that can save costs and does not increase the weight of the overall structure that constitutes the target space. In the case where the target space is specifically inside a vehicle, it can also avoid affecting the vehicle's endurance. By utilizing the characteristics of signal propagation, the disadvantage of the distributed array that cannot directly utilize high-frequency directional characteristics is avoided, so that the window wake-up judgment can be performed only by relying on the distributed array. In particular, the distributed microphone array can be a distributed arrangement of microphones in the space where the sound source is located, and the distance between the sound source and the microphones is of similar magnitude. At the same time, it does not rely on other sensors such as images, and only uses the microphone signal to determine whether the specific wake-up is outside the target space. First, the voice information collected by the microphone device is obtained, and the voice information is recognized. If the voice information contains keywords that wake up the smart device in the target space, the audio signal corresponding to the voice information is parsed to obtain the sound source directionality information. Then, based on the sound source directionality information, it is determined whether the audio signal is multi-path propagation. If the audio signal is multi-path propagation, it is determined that the wake-up behavior corresponding to the voice information is the behavior of waking up the smart device in the target space from outside the target space. By applying the technical solutions of the embodiments of the present application, it is possible to not rely on other sensors such as images, the cost is low, and distributed arrays are supported. There are no strict requirements for array type and layout, and the processing is simple and the accuracy is high.

[0144] FIG5 shows a schematic diagram of a process provided in an embodiment of the present application.

[0145] The wake-up module is used to identify the audio signal within the time period when the keyword appears in the audio signal. The wake-up module inputs the voice signal of the time period when the wake-up word appears into the target space outside wake-up judgment module; the judgment module calculates the data covariance matrix based on the obtained time period information, specifically by calculating the frequency point signal-to-noise ratio, and selects the target frequency point with a higher signal-to-noise ratio. For the target frequency point, the noise data is estimated by the noise energy estimation module, and the estimated signal covariance matrix corresponding to the target frequency point is obtained after removing the noise data; then the estimated signal covariance matrix is ​​eigendecomposed and the main eigenvalue concentration is calculated, and it is determined whether it is a wake-up outside the target space by comparing it with the threshold. If not, the wake-up result is accepted and the smart device responds. Specifically, when the target space is inside the car, the wake-up module is used to identify the audio signal within the time period when the keyword appears in the audio signal. The wake-up module inputs the voice signal of the time period when the wake-up word appears into the outside-car wake-up judgment module; the judgment module calculates the data covariance matrix through the obtained time period information, specifically by calculating the frequency point signal-to-noise ratio, and selects the target frequency point with a higher signal-to-noise ratio. For the target frequency point, the noise data is estimated by the noise energy estimation module, and the estimated signal covariance matrix corresponding to the target frequency point is obtained after removing the noise data; then the estimated signal covariance matrix is ​​eigendecomposed and the main eigenvalue concentration is calculated, and it is determined whether it is an outside-car wake-up by comparing with the threshold. If not, the wake-up result is accepted and the smart device responds.

[0146] By applying the technical solution of this embodiment, it is possible to accurately determine whether the wake-up of the smart device in the target space comes from outside the target space, and then accurately perform corresponding suppression control. In addition, this technical solution does not require the addition of sound insulation materials outside the target space, which can save costs. When the target space is specifically the space inside the vehicle, it will not increase the weight of the vehicle in the target space, thereby avoiding affecting the endurance of the vehicle in the target space.

[0147] This embodiment provides a method for determining a sound source position, as shown in FIG6 , the method including:

[0148] Step 301: Acquire voice information collected by a microphone device and perform voice recognition on the voice information.

[0149] Among them, the microphone device may include a distributed microphone array, and these microphones are dispersedly arranged in the space where the sound source is located. The distance between the sound source and the microphones and the distance between the microphones are at a similar order of magnitude. It should be noted that this embodiment has no restrictions on the arrangement of the microphone device.

[0150] Microphone devices have the function of collecting voice information; voice information may include voice information within the target space and voice information outside the target space, such as voice information inside the car and voice information outside the car, and voice information indoors and outdoors; voice recognition is a high-tech technology that allows machines to convert voice signals into corresponding text or commands through recognition and understanding. It is widely used in various fields such as industry, home appliances, communications, automotive electronics, medical care, home services, consumer electronics, etc.

[0151] In a specific application scenario, when obtaining voice information collected by a microphone device and performing voice recognition on the voice information, the method of this embodiment may include: using a communication device to obtain the voice information collected by the microphone device, and using a voice recognition tool to perform voice recognition on the voice information. This embodiment does not limit this.

[0152] In the embodiment of the present application, the target space includes indoors, inside a vehicle, etc. Specifically, the target space may be a cabin space inside a vehicle.

[0153] Step 302: Analyze the audio signal corresponding to the voice information to obtain sound source directionality information.

[0154] In the embodiments of the present application, the directivity of a sound source can be the response of a transducer in a specified plane passing through the center of the sound as a function of the direction of the emitted or incident sound wave, at a fixed frequency. When measuring the sound pressure level radiated by a sound source at a fixed distance from the sound source within a given frequency band, it is often found that the sound pressure level varies in different directions from the sound source. The low-frequency radiation of many noise sources is almost non-directional, but its directivity increases with increasing frequency. This is because the sound waves radiated from different parts of the vibration source arrive at different points in space at different times, resulting in interference and uneven directional radiation.

[0155] In the embodiment of the present application, the smart device (such as a voice assistant) can be a smart device on a vehicle or a smart device indoors. When the voice information is determined to be voice information within the target space, it can be determined that the voice information is information for executing a corresponding voice command using the smart device.

[0156] In-vehicle smart devices (such as voice assistants) can refer to intelligent voice systems that are designed to avoid distractions caused by prolonged screen reading. These systems, combined with voice assistant services, can execute simple commands, significantly improving driving safety. In specific application scenarios, if the voice information is recognized as being from within the vehicle, the information is determined to be information used by the smart device to execute a corresponding voice command to the vehicle.

[0157] Smart devices (such as voice assistants) can also be indoor smart devices (such as voice assistants). Indoor smart devices (such as voice assistants) can refer to intelligent voice systems that are combined with voice assistant services to facilitate the control of other indoor devices and can execute simple commands, greatly improving the convenience of device control. In a specific application scenario, if the voice information is recognized as voice information in the car, it is determined that the voice information is information for executing corresponding voice commands on other indoor devices using the smart device.

[0158] In an embodiment of the present application, for voice information inside the vehicle, due to the low reflection inside the vehicle, the microphone device mainly receives the audio signal of a single direct path. However, for voice information outside the vehicle, due to the obstruction of the vehicle body and windows, no direct path can be formed. The sound can only be transmitted into the vehicle through the vibration of solid objects, thereby transforming the point sound source into a surface sound source, resulting in a multipath propagation effect. Based on the two different propagation situations, this embodiment utilizes the characteristics of voice signal propagation to circumvent the disadvantage that the distributed array in the microphone device cannot directly utilize the high-frequency directional characteristics. It can also make decisions about voice outside the window by relying solely on the distributed array. That is, it analyzes the audio signal, and then determines the directionality of the sound source based on the sound frequency and direction. The sound directionality is used to determine whether the audio signal is multipath propagation, and then distinguish whether the voice information comes from inside or outside the vehicle. It is used to determine whether the voice information is a user instruction inside the vehicle. Smart devices (such as voice assistants) can determine whether to execute the corresponding operation of the voice information based on the above analysis.

[0159] For example, for indoor voice information, due to the small indoor reflection, the indoor microphone device mainly receives the audio signal of a single direct path; while for outdoor voice information, due to the obstruction of walls, doors and windows, no direct path can be formed, and the sound can only be transmitted into the room through the vibration of solid objects, thereby changing the point sound source into a surface sound source, resulting in a multi-path propagation effect. Based on the two different propagation situations, this embodiment utilizes the characteristics of voice signal propagation to avoid the disadvantage that the distributed array in the microphone device cannot directly utilize the high-frequency directional characteristics. It can also make the judgment of the voice outside the window relying only on the distributed array, that is, based on the audio signal, the directionality of the sound source is determined according to the sound frequency and direction. The sound directionality is used to determine whether the audio signal is multi-path propagation, and then distinguish whether the voice information comes from indoors or outdoors. It is used to determine whether the voice information is an instruction from the user indoors. Smart devices (such as voice assistants) can determine whether to execute the operation corresponding to the voice information based on the above analysis.

[0160] Step 303: Determine the sound source position of the audio signal based on the sound source directivity information.

[0161] In some examples, the audio signal may include propagation through a sound propagation medium within the target space (such as inside a car or indoors) and propagation through a sound propagation medium outside the target space. The sound propagation medium within the target space may be air propagation, and the sound propagation medium outside the target space may be window glass or a wall of the target space (or a structural shell constituting the target space) as well as air propagation, etc. Multipath propagation may be sound propagation along multiple paths formed by voice information outside the target space. In the case where the target space is inside a car, the voice information outside the car may be command information sent by a person outside the car to an intelligent device inside the car (such as a voice assistant) when the car is stopped, the windows and doors are closed, and the car system is powered on. This embodiment does not limit this.

[0162] For an embodiment of the present application, when judging whether an audio signal is multi-path propagated based on the sound source directionality information, the method of this embodiment may include: judging whether the audio signal is multi-path propagated based on the sound source directionality information through an intelligent device (such as a voice assistant) or a vehicle host or a server, etc., and this embodiment does not limit this.

[0163] In the embodiment of the present application, if the audio signal propagates via multiple paths, the audio signal is determined to be speech information outside the target space, i.e., the sound source location of the audio signal is outside the target space. If the audio signal propagates via a single direct path, the audio signal is determined to be speech information within the target space, i.e., the sound source location of the audio signal is within the target space.

[0164] For example, if the above-mentioned recognition and analysis determine that the audio signal is multi-path propagation, then the audio information can be determined to be voice information outside the target space. For example, at this time, the smart device (such as a voice assistant) can be controlled not to execute the instructions corresponding to the audio signal.

[0165] In an embodiment of the present application, if the target space includes the interior of a vehicle and the audio signal is propagated through multiple paths, the process of determining that the audio signal is voice information outside the target space and controlling the smart device not to execute the instruction corresponding to the audio signal includes: if the audio signal is propagated through multiple paths, determining that the audio signal is an audio signal outside the vehicle and controlling the smart device inside the vehicle not to execute the instruction corresponding to the audio signal; if the audio signal is propagated through a single direct path, determining that the audio signal is an audio signal inside the vehicle and controlling the smart device inside the vehicle to execute the instruction corresponding to the audio signal. If the target space includes an indoor space and the audio signal is propagated through multiple paths, determining that the audio signal is an audio signal outside the vehicle and controlling the smart device inside the vehicle not to execute the instruction corresponding to the audio signal; if the audio signal is propagated through a single direct path, determining that the audio signal is an audio signal inside the vehicle and controlling the smart device inside the vehicle to execute the instruction corresponding to the audio signal.

[0166] The embodiment of the present application provides a method for determining the position of a sound source, which can determine the position of the sound source, and only use the smart device to execute the instructions corresponding to the audio signal in the target space, and do not use the smart device to execute the instructions corresponding to the audio signal outside the target space, thereby improving the accuracy of using the smart device to operate other devices in the target space. Specifically, first obtain the voice information collected by the microphone device, and perform voice recognition on the voice information; parse the audio signal corresponding to the voice information to obtain the sound source directionality information; and then determine the sound source position of the audio signal based on the sound source directionality information. If the audio signal is multi-path propagated, it is determined that the voice information is voice information outside the target space, and the smart device in the target space is not used to execute the instructions corresponding to the voice information. By applying the technical solution of the embodiment of the present application, it is possible to accurately use smart devices to operate other devices in the target space.

[0167] Furthermore, in order to facilitate understanding of the implementation process of this embodiment, the principles involved in this embodiment are specifically described below. It should be noted that the following example content is only one optional method of the method of this embodiment, and this embodiment does not limit the specific implementation method of the method.

[0168] Regarding speech information, it can be divided into speech information within the target space and speech information outside the target space.

[0169] For the speech information in the target space, since the reflection in the target space is small, the microphone mainly receives the audio signal of the direct path. The data model can be expressed as follows: x(t) = d θ *s(t)+n(t) (Formula 1)

[0170] In formula 1, x(t) represents the microphone received signal vector, d θ The steering vector represents the sound source position, s(t) represents the sound source signal, and n(t) represents the noise signal vector.

[0171] Furthermore, the data covariance matrix obtained for the speech information in the target space can be expressed as the following formula 2:

[0172] In formula 2, R represents the covariance matrix in the single-path propagation scenario, that is, the data covariance matrix obtained by the voice information in the target space, E represents the mean operation, x(t) represents the microphone received signal vector, d θ The steering vector representing the sound source position, σ 2 represents the noise power, H represents the conjugate transpose operation, and I represents the identity matrix.

[0173] In specific application scenarios, for voice information outside the target space, due to the obstruction of the overall structure of the target space (in the case of a car, the overall structure of the target space can be the car body) and the windows of the target space, no direct path can be formed. The sound can only be transmitted into the target space through the vibration of the solid, thus changing the point sound source into a surface sound source, resulting in a multipath propagation effect. Assuming that there are P propagation paths, the data model can be expressed as follows:

[0174] In formula 3, xc(t) represents the microphone received signal vector in the multipath propagation scenario, that is, the data covariance matrix obtained by the voice information outside the target space, d θ The steering vector represents the sound source position, s(t) represents the sound source signal, and n(t) represents the noise signal vector.

[0175] Furthermore, the data covariance matrix obtained for the speech information outside the target space can be expressed as the following formula 4:

[0176] In formula 4, R c represents the covariance matrix in the multipath propagation scenario, E represents the mean operation, x c (t) represents the microphone received signal vector in the multipath propagation scenario, The steering vector representing the sound source position, σ 2 represents the noise power, H represents the conjugate transpose operation, and I represents the identity matrix.

[0177] In the above formula 4 and formula 5, there may be noise data. In this embodiment, the noise data may be removed to obtain an estimated signal covariance matrix.

[0178] For example, the estimated signal covariance matrix of the speech information in the target space can be expressed as follows:

[0179] In formula 5, R s Represents the estimated signal covariance matrix of the speech information in the target space, R represents the covariance matrix, σ 2 represents the noise power, I represents the identity matrix, d θ represents the steering vector of the sound source position, and H represents the conjugate transpose operation.

[0180] The estimated signal covariance matrix of the speech information outside the target space can be expressed as follows:

[0181] In formula six, The estimated signal covariance matrix, R, represents the speech information outside the target space. c represents the covariance matrix, σ2 represents the noise power, I represents the identity matrix, represents the steering vector of the sound source position, and H represents the conjugate transpose operation.

[0182] Based on Formula 5 and Formula 6, it can be analyzed that since the voice information in the target space is transmitted in a direct path, the obtained signal covariance matrix is ​​a rank-one matrix, while the signal covariance matrix rank of the voice information outside the target space is greater than 1. Therefore, this feature can be used to identify the voice information in the target space and the voice information outside the target space.

[0183] According to the above exemplary principles, the sound source determination method provided in this embodiment includes the following steps:

[0184] Step 401: Acquire voice information collected by a microphone device and perform voice recognition on the voice information.

[0185] For the specific implementation process of the embodiment of the present application, please refer to the relevant description in step 101 of the embodiment, which will not be repeated here.

[0186] Step 402: Calculate and obtain a data covariance matrix based on the audio signal.

[0187] In the embodiment of the present application, the data covariance matrix may be a covariance matrix obtained from audio signal data.

[0188] Optionally, step 402 of the embodiment may specifically include: extracting signal data of the audio signal; and calculating the signal data to obtain a data covariance matrix. In this embodiment, regardless of whether the speech information is within the target space or outside the target space, a data covariance matrix is ​​calculated based on the audio information of the speech information, such as in the form of Formula 2 or Formula 4 above. Whether the speech signal is within the target space or outside the target space can be determined based on further analysis results, such as by executing the process shown in steps 403 to 405.

[0189] Step 403: Perform noise removal on the data covariance matrix to obtain an estimated signal covariance matrix.

[0190] Optionally, step 403 may specifically include: subtracting a noise covariance matrix corresponding to the estimated noise energy of the audio signal from the data covariance matrix to obtain an estimated signal covariance matrix.

[0191] This embodiment can estimate noise energy and thereby obtain noise data in the audio signal, such as a noise covariance matrix corresponding to the estimated noise energy of the audio signal. The data covariance matrix is ​​then subjected to a noise energy removal process using the estimated noise energy. For example, the noise covariance matrix corresponding to the estimated noise energy of the audio signal is subtracted from the data covariance matrix to obtain an estimated signal covariance matrix. The specific form can be as shown in Formula 5 or Formula 6 above.

[0192] Step 404: perform eigendecomposition on the estimated signal covariance matrix and calculate the main eigenvalue concentration, and use the main eigenvalue concentration as the sound source directivity information.

[0193] Optionally, step 404 may specifically include: first calculating the audio signal to obtain the signal-to-noise ratio of different frequency points; then selecting a target frequency point whose signal-to-noise ratio is greater than a preset threshold among the different frequency points; then performing eigendecomposition on the estimated signal covariance matrix corresponding to the target frequency point, and then calculating the main eigenvalue concentration based on the matrix eigenvalues ​​obtained by the eigendecomposition.

[0194] In specific application scenarios, noise estimation can exhibit certain deviations. If a frequency point with a relatively low signal-to-noise ratio is selected to obtain the corresponding estimated signal covariance matrix, the corresponding principal eigenvalue will be unstable when the matrix is ​​subjected to eigendecomposition, affecting the accuracy of subsequent evaluations. Therefore, this embodiment selects a frequency point with a signal-to-noise ratio greater than a certain threshold to obtain the corresponding estimated signal covariance matrix. When the matrix is ​​subjected to eigendecomposition, the corresponding principal eigenvalue is relatively stable. This allows accurate judgment of speech information within and outside the target space based on the concentration of the principal eigenvalue.

[0195] When the estimated signal covariance matrix is ​​subjected to eigendecomposition, multiple eigenvalues ​​λ may be obtained, and the largest eigenvalue among them is selected as the main eigenvalue.

[0196] For example, the main eigenvalue concentration is obtained by using the following formula 7:

[0197] In Formula 7, ε is the main eigenvalue concentration, λ1 is the main eigenvalue, M is the number of matrix eigenvalues, and λi is the i-th eigenvalue.

[0198] Step 405: Compare the main eigenvalue concentration with a preset threshold value. If the main eigenvalue concentration is less than the preset threshold value, it is determined that the audio signal is multipath propagated.

[0199] In the embodiment of the present application, the absolute value obtained by subtracting the main eigenvalue concentration from 1 is compared with a preset threshold value to determine a reference value, and then the reference value is used to determine whether the audio signal is single-path propagation or multi-path propagation.

[0200] Optionally, through step 404 of the above embodiment, it can be determined that when the speech information is speech information within the target space, the main eigenvalue concentration is close to 1, while when the speech information is speech information outside the target space, it is significantly less than 1, thereby distinguishing whether the speech information is outside the target space. A threshold value is preset, the main eigenvalue concentration is subtracted from 1, and an absolute value analysis is performed to obtain a concentration difference. If the concentration difference is less than the preset threshold value, the speech information is speech information within the target space; if the concentration difference is greater than the preset threshold value, the speech information is speech information outside the target space. The threshold value is determined by performing absolute value processing based on the difference between the speech signal within the target space and 1, and the difference between the speech information outside the target space and 1, and then performing absolute value processing, and then performing comprehensive statistics on the absolute values ​​corresponding to the speech information within the target space and the absolute values ​​corresponding to the speech information outside the target space.

[0201] For the embodiment of the present application, in the case of voice information inside the car, the main eigenvalue concentration is close to 1, while in the case of voice information outside the car, it is significantly less than 1, so that it can be distinguished whether it is voice information outside the car. A threshold value is preset, the main eigenvalue concentration is subtracted from 1, and an absolute value analysis is performed to obtain the concentration difference. If the concentration difference is less than the preset threshold value, the voice information is voice information inside the car; if the concentration difference is greater than the preset threshold value, the voice information is voice information outside the car. Among them, the threshold value is determined by subtracting the difference between the voice information inside the car and 1, and then performing absolute value processing; and subtracting the difference between the voice information outside the car and 1, and then performing absolute value processing, and then comprehensively statistically combining the absolute values ​​corresponding to the above-mentioned voice information inside the car and the absolute values ​​corresponding to the voice information outside the car to determine a threshold value. In the case of voice information indoors, the main eigenvalue concentration is close to 1, while in the case of voice information outdoors, it is significantly less than 1, so that it can be distinguished whether it is voice information outdoors. A threshold is preset, and the main eigenvalue concentration is subtracted from 1. An absolute value analysis is performed to determine the concentration difference. If the concentration difference is less than the preset threshold, the voice information is indoors; if the concentration difference is greater than the preset threshold, the voice information is outdoor. The threshold is determined by subtracting the difference between the indoor voice information and 1, then performing absolute value processing; and by subtracting the difference between the outdoor voice information and 1, then performing absolute value processing. The absolute values ​​corresponding to the indoor and outdoor voice information are then statistically combined to determine the threshold.

[0202] For example, statistics are collected for the absolute value A corresponding to the subtraction of 1 from the semantic information inside a vehicle (or indoors) and the absolute value B corresponding to the subtraction of 1 from the speech information outside a vehicle (or outdoors). Absolute values ​​A include 0.3, 0.5, 0.2, 0.4, and 0.3, and absolute values ​​B include 0.6, 0.7, 0.5, 0.8, and 0.9. The average values ​​of absolute values ​​A and B are 0.34 and 0.7, respectively. Furthermore, as shown in the above embodiment, the concentration of the main eigenvalues ​​of the speech information inside a vehicle (or indoors) is close to 1, while the concentration of the main eigenvalues ​​of the speech information outside a vehicle (or outdoors) is significantly less than 1. The average value of absolute values ​​A, 0.34, is less than the average value of absolute values ​​B, 0.7. Therefore, the above values ​​are valid. The quotient of 0.52, obtained by adding the average value of absolute values ​​A, 0.34, and the average value of absolute values ​​B, 0.7, and dividing by 2, is used as the threshold value. The current main eigenvalue concentration TZ is extracted as 0.45 and subtracted from 1. The absolute value of the difference obtained is 0.55, which is the concentration difference. The concentration difference 0.55 is greater than the threshold value 0.52, which determines that the audio signal is multipath propagation.

[0203] Through the specific implementation of the above embodiment, it is possible to determine whether the voice information is transmitted through a single path or multiple paths. The single path is the voice information within the target space, while the multipath is the voice information outside the target space. Furthermore, the voice information can be determined to be within the target space or outside the target space, thereby performing relative processing on the voice information, thereby avoiding the impact on user experience and safety risks caused by incorrect operations. At the same time, voice control failures caused by the low voice of the person in the target space issuing a voice command to the smart device will not be caused.

[0204] Step 406: If the audio signal propagates via multiple paths, the voice information is determined to be voice information outside the target space; if the audio signal propagates via a single direct path, the voice information is determined to be voice information within the target space.

[0205] In some application scenarios, embodiments of the present application may further include: if the voice information is determined to be within a target space, controlling the smart device to execute a command corresponding to the voice information. For example, if the target space is within a vehicle, the voice information is within the target space, and the command corresponding to the voice information is to open a vehicle door, then the smart device executes the command corresponding to the voice information to open the vehicle door.

[0206] Embodiments of the present application provide a method for determining the location of a sound source. This method can determine the location of a sound source and, in the case of a target space specifically within a vehicle, prevent voice information outside the vehicle from controlling the vehicle. By leveraging the characteristics of signal propagation, the method overcomes the disadvantage of distributed arrays, which cannot directly utilize high-frequency directional characteristics. This method allows the location of sound sources outside a window to be determined solely using a distributed array. The distributed microphone array can be configured such that the microphones are dispersed throughout the space where the sound source resides, with the distance between the sound source and the microphones being of similar magnitude to the distance between the microphones. Furthermore, the method does not rely on other sensors, such as imaging, but only uses the microphone signals to determine whether the sound information originates outside the target space. Specifically, the method first acquires voice information collected by the microphone devices and performs speech recognition on the voice information. The audio signal corresponding to the voice information is analyzed to obtain sound source directional information. Based on this sound source directional information, the sound source location of the audio signal is determined. If the audio signal propagates through multiple paths, the voice information is determined to be outside the target space. By applying the technical solutions of the embodiments of the present application, the method eliminates the need for other sensors, such as imaging, resulting in lower costs. Furthermore, the method supports distributed arrays, has no strict requirements on array type and layout, and offers simple processing and high accuracy.

[0207] By applying the technical solution of this embodiment, it is possible to accurately identify voice information as voice information within the target space or voice information outside the target space. For voice information outside the target space, the smart device does not execute the instructions corresponding to the voice information, but only executes the instructions corresponding to the voice information within the target space, thereby accurately performing corresponding suppression control.

[0208] Furthermore, as a specific implementation of the method shown in FIG6 , this embodiment provides a device for determining a sound source position. As shown in FIG7 , the device includes: a first acquiring part 21 , a first analyzing part 22 , and a determining part 23 .

[0209] The first acquisition part 21 is configured to acquire voice information collected by the microphone device and perform voice recognition on the voice information;

[0210] The first analyzing part 22 is configured to analyze the audio signal corresponding to the voice information to obtain sound source directionality information;

[0211] The determination part 23 is configured to determine the sound source position of the audio signal based on the sound source directivity information.

[0212] In a specific application scenario, the first analysis part 22 is specifically configured to calculate a data covariance matrix based on the audio signal; perform noise removal processing on the data covariance matrix to obtain an estimated signal covariance matrix; perform eigendecomposition and calculation on the estimated signal covariance matrix to obtain the main eigenvalue concentration, and use the main eigenvalue concentration as the sound source directionality information.

[0213] In a specific application scenario, the determination part 23 is specifically configured to compare the main eigenvalue concentration with a preset threshold value; if the main eigenvalue concentration is less than the preset threshold value, the speech information is determined to be speech information outside the target space.

[0214] In a specific application scenario, the first analyzing part 22 is specifically configured to extract signal data corresponding to the audio signal; and calculate the signal data to obtain a data covariance matrix.

[0215] In a specific application scenario, the first analyzing part 22 is specifically configured to subtract the noise covariance matrix corresponding to the estimated noise energy of the audio signal from the data covariance matrix to obtain an estimated signal covariance matrix.

[0216] In a specific application scenario, the first analysis part 22 is specifically configured to calculate the audio signal to obtain the signal-to-noise ratio of different frequency points; select a target frequency point whose signal-to-noise ratio is greater than a preset threshold among the different frequency points; perform eigendecomposition on the signal covariance matrix estimated corresponding to the target frequency point; and calculate the main eigenvalue concentration based on the matrix eigenvalue obtained by the eigendecomposition.

[0217] It should be noted that, for other corresponding descriptions of the functional units involved in the sound source location determination device provided in this embodiment, reference can be made to the corresponding description in FIG6 , which will not be repeated here.

[0218] Based on the method shown in FIG. 6 , this embodiment further provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by the second processor, the method shown in FIG. 6 is implemented.

[0219] Based on this understanding, the technical solution of this embodiment can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, USB flash drive, mobile hard disk, etc.), and includes a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods of each implementation scenario of this embodiment.

[0220] Based on the above-mentioned method shown in Figure 6 and the virtual device embodiment shown in Figure 7, in order to achieve the above-mentioned purpose, the embodiment of the present application also provides an electronic device that can be configured on the vehicle (such as a new energy vehicle or a traditional vehicle) side, and the device includes a storage medium and a second processor; the storage medium is used to store a computer program; the second processor is used to execute the computer program to implement the above-mentioned method shown in Figure 6.

[0221] Optionally, the physical device may further include a user interface, a network interface, a camera, a radio frequency (RF) circuit, a sensor, an audio circuit, a Wi-Fi module, and the like. The user interface may include a display, an input unit such as a keyboard, and the like. Optional user interfaces may also include a USB interface, a card reader interface, and the like. The network interface may optionally include a standard wired interface, a wireless interface (such as a Wi-Fi interface), and the like.

[0222] Those skilled in the art will understand that the above-mentioned physical device structure provided in this embodiment does not constitute a limitation on the physical device, and may include more or fewer components, or a combination of certain components, or different component arrangements.

[0223] The storage medium may also include an operating system and a network communication module. The operating system is a program that manages the hardware and software resources of the physical device, supporting the execution of information processing programs and other software and / or programs. The network communication module is used to enable communication between components within the storage medium, as well as with other hardware and software within the physical information processing device.

[0224] Based on the above electronic device, an embodiment of the present application further provides a vehicle, which may specifically include: the sound source location determination device shown in FIG7 , or the above electronic device. The vehicle may specifically be a new energy vehicle or a traditional vehicle.

[0225] Through the description of the above implementation methods, those skilled in the art can clearly understand that the present application can be implemented by means of software plus a necessary general hardware platform, or by hardware.

[0226] Furthermore, as a specific implementation of the method shown in Figures 1 and 4, this embodiment provides a wake-up control device, as shown in Figure 8, which includes: a second acquisition part 31, a second analysis part 32, a judgment part 33, and a control part 34.

[0227] The second acquisition part 31 is configured to acquire voice information collected by the microphone device and perform voice recognition on the voice information;

[0228] The second parsing part 32 is configured to parse the audio signal corresponding to the voice information to obtain sound source directionality information if it is recognized that the voice information contains a keyword for waking up the smart device in the target space;

[0229] The judging part 33 is configured to judge whether the audio signal is multipath propagated based on the sound source directivity information;

[0230] The control part 34 is configured to determine that the wake-up behavior corresponding to the voice information is a behavior of waking up the smart device in the target space outside the target space if the audio signal is multi-path propagated, and control not to wake up the smart device in the target space.

[0231] In a specific application scenario, the second analysis part 32 is specifically configured to calculate a data covariance matrix based on the audio signal; perform noise removal processing on the data covariance matrix to obtain an estimated signal covariance matrix; perform eigendecomposition and calculation on the estimated signal covariance matrix to obtain the main eigenvalue concentration, and use the main eigenvalue concentration as the sound source directionality information.

[0232] In a specific application scenario, the judgment part 33 is specifically configured to compare the main eigenvalue concentration with a preset threshold value; if the main eigenvalue concentration is less than the preset threshold value, it is determined that the audio signal is multipath propagated.

[0233] In a specific application scenario, the second analysis part 32 is further configured to extract signal data of the time period in which the keyword appears in the audio signal; calculate the signal data to obtain a data covariance matrix; and convert the data covariance matrix to obtain a noisy data covariance matrix.

[0234] In a specific application scenario, the second analyzing part 32 is further configured to subtract the noise covariance matrix corresponding to the estimated noise energy of the audio signal from the data covariance matrix to obtain an estimated signal covariance matrix.

[0235] In a specific application scenario, the second analysis part 32 is further configured to calculate the audio signal to obtain the signal-to-noise ratio of different frequency points; select a target frequency point whose signal-to-noise ratio is greater than a preset threshold among the different frequency points; perform eigendecomposition on the estimated signal covariance matrix corresponding to the target frequency point; and calculate the main eigenvalue concentration based on the matrix eigenvalue obtained by the eigendecomposition.

[0236] In a specific application scenario, the judgment part 33 is further configured to determine that the wake-up behavior corresponding to the voice information is the behavior of waking up the smart device in the target space if the audio signal is propagated in a single direct path, and control the wake-up of the smart device in the target space.

[0237] In a specific application scenario, the target space includes the interior of the vehicle.

[0238] The second parsing part 32 is specifically configured to parse the audio signal corresponding to the voice information to obtain sound source directionality information if the voice information contains a keyword for waking up the in-vehicle smart device;

[0239] Accordingly, the control part 34 is specifically configured to determine that the wake-up behavior corresponding to the voice information is the behavior of waking up the smart device in the car from outside the car if the audio signal is multi-path propagated, and control not to wake up the smart device in the car.

[0240] In a specific application scenario, the target space includes indoors.

[0241] The second parsing part 32 is specifically configured to parse the audio signal corresponding to the voice information to obtain sound source directionality information if it is recognized that the voice information contains a keyword for waking up the indoor smart device;

[0242] Accordingly, the control part 34 is specifically configured to determine that the wake-up behavior corresponding to the voice information is an outdoor wake-up behavior of the indoor smart device if the audio signal is multi-path propagated, and control not to wake up the indoor smart device.

[0243] It should be noted that, for other corresponding descriptions of the functional units involved in the wake-up control device provided in this embodiment, reference may be made to the corresponding descriptions in FIG. 1 and FIG. 4 , which will not be repeated here.

[0244] Based on the above methods shown in FIG. 1 and FIG. 4 , this embodiment also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by the first processor, the above methods shown in FIG. 1 and FIG. 4 are implemented.

[0245] Based on this understanding, the technical solution of this embodiment can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, USB flash drive, mobile hard disk, etc.), and includes a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods of each implementation scenario of this embodiment.

[0246] Based on the above-mentioned method shown in Figures 1 and 4, and the virtual device embodiment shown in Figure 8, in order to achieve the above-mentioned purpose, the embodiment of the present application also provides an electronic device that can be configured on the vehicle (such as a new energy vehicle or a traditional vehicle) side, and the device includes a storage medium and a first processor; the storage medium is used to store a computer program; the first processor is used to execute the computer program to implement the above-mentioned method shown in Figures 1 and 4.

[0247] Optionally, the physical device may further include a user interface, a network interface, a camera, a radio frequency (RF) circuit, a sensor, an audio circuit, a Wi-Fi module, and the like. The user interface may include a display, an input unit such as a keyboard, and the like. Optional user interfaces may also include a USB interface, a card reader interface, and the like. The network interface may optionally include a standard wired interface, a wireless interface (such as a Wi-Fi interface), and the like.

[0248] Those skilled in the art will understand that the above-mentioned physical device structure provided in this embodiment does not constitute a limitation on the physical device, and may include more or fewer components, or a combination of certain components, or different component arrangements.

[0249] The storage medium may also include an operating system and a network communication module. The operating system is a program that manages the hardware and software resources of the physical device, supporting the execution of information processing programs and other software and / or programs. The network communication module is used to enable communication between components within the storage medium, as well as with other hardware and software within the physical information processing device.

[0250] Based on the above electronic device, an embodiment of the present application further provides a vehicle, which may specifically include: the wake-up control device shown in FIG8 , or the above electronic device. The vehicle may specifically be a new energy vehicle or a traditional vehicle.

[0251] Through the description of the above implementation methods, those skilled in the art can clearly understand that the present application can be implemented by means of software plus the necessary general hardware platform, or by means of hardware. By applying the solution of this embodiment, compared with the currently existing wake-up control method, the present application provides a wake-up control method, a sound source location determination method, a device, an electronic device, and a vehicle. When the target space is inside the vehicle, it can be determined whether the wake-up comes from outside the vehicle only through the microphone array inside the vehicle when the vehicle is stopped, the doors and windows are closed, and the in-vehicle system is powered on. If so, it is suppressed. This can avoid the limited sound insulation effect of the vehicle body, and increasing the sound insulation will further increase the hardware cost and reduce the battery life; avoid the poor wake-up suppression effect for loud sound sources in parking spaces close to the windows; and reduce the existing method of locating the sound source through a small-pitch microphone array inside the vehicle to determine whether it is outside the vehicle. The restrictions on the arrangement of microphones inside the vehicle are reduced. Specifically, the method first obtains the voice information collected by the microphone device and performs voice recognition on the voice information; if the voice information is recognized to contain keywords that wake up the smart device in the car, the audio signal corresponding to the voice information is parsed to obtain the sound source directionality information; then, based on the sound source directionality information, it is determined whether the audio signal is multi-path propagated; if the audio signal is multi-path propagated, it is determined that the wake-up behavior corresponding to the voice information is the behavior of waking up the smart device in the car from outside the car. By applying the technical solution of this application, it is possible to not rely on other sensors such as images, the cost is low, and it supports distributed arrays. There are no strict requirements for array type and layout, and the processing is simple and the accuracy is high.

[0252] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprises" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device that includes a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a..." do not exclude the presence of other identical elements in the process, method, article or device that includes the elements.

[0253] The foregoing is merely a list of specific embodiments of the present application, intended to enable those skilled in the art to understand and implement the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application is not limited to the embodiments described herein, but is intended to conform to the broadest scope consistent with the principles and novel features of the present application. Industrial Applicability

[0254] The embodiments of the present application provide a wake-up control method, a sound source location determination method, an apparatus, an electronic device, and a vehicle. The wake-up control method includes: obtaining voice information collected by a microphone device, and performing voice recognition on the voice information; if the voice information is identified to contain keywords for waking up smart devices in the target space, parsing the audio signal corresponding to the voice information to obtain sound source directionality information; then, based on the sound source directionality information, determining whether the audio signal is multi-path propagated; if the audio signal is multi-path propagated, determining that the wake-up behavior corresponding to the voice information is a behavior of waking up smart devices in the target space from outside the target space, and controlling not to wake up the smart devices in the target space. The implementation scheme using the above method can accurately determine whether the wake-up of smart devices in the target space comes from outside the target space, and then accurately perform corresponding suppression control. In addition, the technical scheme does not require the addition of additional sound insulation materials in the target space, which can save costs.

Claims

1. A wake-up control method, the method comprising: Acquire voice information collected by the microphone device, and perform voice recognition on the voice information; If it is recognized that the voice information contains a keyword for waking up the smart device in the target space, the audio signal corresponding to the voice information is parsed to obtain the sound source directionality information; Based on the sound source directivity information, determining whether the audio signal is multipath propagated; If the audio signal is multi-path propagated, it is determined that the wake-up behavior corresponding to the voice information is a behavior of waking up the smart device in the target space outside the target space, and the smart device in the target space is controlled not to be woken up.

2. The method according to claim 1, wherein: The step of parsing the audio signal corresponding to the voice information to obtain the sound source directivity information includes: Calculating a data covariance matrix according to the audio signal; Performing noise removal on the data covariance matrix to obtain an estimated signal covariance matrix; Performing eigendecomposition and calculation on the estimated signal covariance matrix to obtain a main eigenvalue concentration, and using the main eigenvalue concentration as the sound source directivity information.

3. The method according to claim 2, wherein: The determining, based on the sound source directivity information, whether the audio signal is multipath propagated includes: Comparing the main eigenvalue concentration with a preset threshold value; If the main eigenvalue concentration is less than the preset threshold value, it is determined that the audio signal is multipath propagated.

4. The method according to claim 2, wherein: The step of calculating a data covariance matrix based on the audio signal includes: Extracting signal data of a time period in which the keyword appears in the audio signal; The signal data is calculated to obtain a data covariance matrix.

5. The method according to claim 2, wherein: The step of performing noise removal processing on the data covariance matrix to obtain an estimated signal covariance matrix includes: A noise covariance matrix corresponding to the estimated noise energy of the audio signal is subtracted from the data covariance matrix to obtain an estimated signal covariance matrix.

6. The method according to claim 2, wherein: The step of performing eigendecomposition and calculating the estimated signal covariance matrix to obtain the main eigenvalue concentration includes: Calculating the audio signal to obtain signal-to-noise ratios at different frequencies; Selecting a target frequency point whose signal-to-noise ratio is greater than a preset threshold value among the different frequency points; Performing eigendecomposition on a signal covariance matrix estimated corresponding to the target frequency point; Based on the matrix eigenvalues ​​obtained by eigendecomposition, the main eigenvalue concentration is calculated.

7. The method according to any one of claims 1 to 6, wherein: After determining whether the audio signal is multipath propagated based on the sound source directivity information, the method further includes: If the audio signal is transmitted via a single direct path, it is determined that the wake-up behavior corresponding to the voice information is the target The smart device in the target space is awakened within the target space, and the smart device in the target space is awakened by the control.

8. The method according to claim 1, wherein: The target space includes the interior of a car. If it is recognized that the voice information contains a keyword for waking up a smart device in the target space, an audio signal corresponding to the voice information is parsed to obtain sound source directivity information, including: If it is recognized that the voice information contains a keyword for waking up the in-vehicle smart device, the audio signal corresponding to the voice information is parsed to obtain the sound source directionality information; Correspondingly, if the audio signal is multipath propagated, determining that the wake-up behavior corresponding to the voice information is a behavior of waking up the smart device in the target space outside the target space, and controlling not to wake up the smart device in the target space, includes: If the audio signal is multi-path propagated, it is determined that the wake-up behavior corresponding to the voice information is a behavior of waking up the in-vehicle smart device outside the vehicle, and the in-vehicle smart device is controlled not to be woken up.

9. The method according to claim 1, wherein: The target space includes indoors. If it is recognized that the voice information contains a keyword for waking up a smart device in the target space, an audio signal corresponding to the voice information is parsed to obtain sound source directivity information, including: If it is recognized that the voice information contains a keyword for waking up the indoor smart device, the audio signal corresponding to the voice information is parsed to obtain the sound source directionality information; Correspondingly, if the audio signal is multipath propagated, determining that the wake-up behavior corresponding to the voice information is a behavior of waking up the smart device in the target space outside the target space, and controlling not to wake up the smart device in the target space, includes: If the audio signal is multi-path propagated, it is determined that the wake-up behavior corresponding to the voice information is an outdoor wake-up behavior of the indoor smart device, and the indoor smart device is controlled not to be woken up.

10. A method for determining a sound source position, comprising: Acquire voice information collected by the microphone device, and perform voice recognition on the voice information; Analyze the audio signal corresponding to the voice information to obtain sound source directionality information; Based on the sound source directivity information, a sound source position of the audio signal is determined.

11. A device for determining a sound source position, comprising: The first acquisition part is configured to acquire the voice information collected by the microphone device and perform voice recognition on the voice information; A first analysis part is configured to analyze the audio signal corresponding to the voice information to obtain sound source directivity information; The determination part is configured to determine the sound source position of the audio signal based on the sound source directivity information.

12. A wake-up control device, comprising: The second acquisition part is configured to acquire the voice information collected by the microphone device and perform voice recognition on the voice information; The second parsing part is configured to parse the audio signal corresponding to the voice information to obtain the sound source directivity information if it is recognized that the voice information contains a keyword for waking up the smart device in the target space; A judging part, configured to judge whether the audio signal is multipath propagated based on the sound source directivity information; The control part is configured to determine that the wake-up behavior corresponding to the voice information is a behavior of waking up the smart device in the target space outside the target space if the audio signal is multi-path propagated, and control not to wake up the smart device in the target space.

13. A computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a first processor, the method according to any one of claims 1 to 9 is implemented; and when the computer program is executed by a second processor, the method according to claim 10 is implemented.

14. An electronic device, comprising a storage medium, a first processor, a second processor, a computer program stored on the storage medium and executable on the first processor, and a computer program stored on the storage medium and executable on the second processor, wherein the first processor implements the method described in any one of claims 1 to 9 when executing the computer program, and the second processor implements the method described in claim 10 when executing the computer program.

15. A vehicle comprising: The device according to claims 11-12, or the electronic device according to claim 14.

Images