Sound event detection using surface mounted vibration sensor

By installing a vibration sensor on the surface of the vehicle, sensing vibration caused by the sound field and processing data, the problem of vulnerability of traditional acoustic microphones in harsh environments is solved, and more reliable acoustic event detection is achieved.

CN118922694APending Publication Date: 2024-11-08HARMAN INT IND INC
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202280094059.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2022-03-30
Publication Date
2024-11-08

AI Technical Summary

Technical Problem

The prior art is difficult to effectively detect the sound field around the vehicle in harsh environments, especially because traditional acoustic microphones are prone to damage and cannot maintain normal operation under external motivations such as water, ice, and dust.

Method used

A surface-mounted vibration sensor is used to induce vibration caused by the incoming sound field through the sound field transducer element of the solid structure, and a computing device is used to process sensor data to identify sound events and provide output data.

Benefits of technology

More reliable and consistent acoustic event detection in harsh environments is achieved, reducing external impact on sensor performance and improving the recognition of emergency sounds and other acoustic events.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118922694A_ABST
    Figure CN118922694A_ABST
Patent Text Reader

Abstract

There is provided a method for detecting a sound field in an external environment of a solid structure, where the solid structure comprises at least one sound transduction element that is part of the solid structure and forms part of an external surface of the solid structure, wherein incoming sound at least partially falls on the sound transduction element. The method includes obtaining sensor data from at least one vibration sensor mounted to a mounting surface of the sound transduction element, where the vibration sensor detects vibrations in the sound transduction element, the vibrations being induced into the sound transduction element by an incoming sound field; processing the sensor data to generate output data; and providing the output data.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] Various examples of the present disclosure generally relate to acoustic event detection using a vibration sensor attached to a surface of a solid structure, such as a vehicle, building, or electronic device. Various examples of the present disclosure specifically relate to acoustic event detection using a vibration sensor mounted to a surface of an acoustic transducing element included in the solid structure and exposed to an acoustic field on an outer surface of the solid structure. Background Art

[0002] With the growth of safety, driver assistance technology and autonomous vehicles, the demand for acoustic and non-acoustic situations and event monitoring around the vehicle has grown widely. Due to the development of these technologies, it has become necessary to detect key sounds from outside the vehicle. Such sounds may include alarms, wake-up words or voice command words from emergency vehicles (e.g., police cars, fire trucks, ambulances), knocks from pedestrians or foreign objects hitting the car body, and other warning acoustic events occurring around the car (e.g., tire explosions, traffic accidents, shootings). For this reason, a given number of sensors need to be integrated in the vehicle to monitor and detect these acoustic events. These sensors must withstand the use of the car in all driving scenarios and are designed to withstand external factors (such as water, ice, dust or debris) that may change the performance of any sensor and endanger the performance of the intended system.

[0003] Acoustic microphones are commonly used to detect sound fields. They rely on a direct air path to sense sound, but are easily damaged in harsh surrounding environments. Summary of the invention

[0004] Therefore, there is a need for advanced techniques for detecting the acoustic field in the surroundings of solid structures. In particular, there is a need for techniques that alleviate or mitigate at least some of the above-mentioned limitations and disadvantages.

[0005] The features of the independent claim meet this need. The features of the dependent claims define further advantageous embodiments.

[0006] In the following, the solution according to the present disclosure is described with respect to the claimed method and with respect to the claimed system, wherein features, advantages or alternative embodiments may be assigned to the other claimed object and vice versa. In other words, the claims relating to the system may be improved with features described in the context of the method, and the method may be improved with features described in the context of the system.

[0007] A method for detecting an acoustic field around a solid structure, such as a vehicle, a building, or an electronic device (e.g., an electronic consumer device), is provided. Such an acoustic field may be generated by acoustic events in a surrounding environment. The surrounding environment may have an atmosphere of a gaseous or liquid medium. The surrounding gaseous or liquid medium may have an interface with the solid structure. Incoming sound may propagate in the surrounding gaseous or liquid medium toward the solid structure and impinge on the solid structure, wherein the incoming sound may be partially reflected and / or partially absorbed, thereby generating vibrations in the solid body of the solid structure. It should be understood that the disclosed technology may be applied to any kind of solid structure that may be surrounded by a sound-bearing medium (e.g., air or water) through which the incoming sound may propagate and impinge on the solid structure.

[0008] The solid structure comprises at least one sound transducing element which is part of the solid structure and forms at least a part of an outer surface of the solid structure, wherein the sound field at least partially impinges on the sound transducing element.

[0009] In one step, sensor data is obtained from at least one vibration sensor mounted to a mounting surface of the sound transducing element, wherein the vibration sensor detects vibrations in the sound transducing element, the vibrations being induced into the sound transducing element by an incoming sound field. In another step, the sensor data is processed to generate output data, and in a further step, the output data is provided.

[0010] Additionally, a system is provided that is configured for detecting an acoustic field incident upon a solid structure, such as, for example, a vehicle, a building, or an electronic device.

[0011] The solid structure includes at least one sound transducer element formed of a solid material, the at least one sound transducer element may be part of the solid structure and may have an outer surface on which the sound field may fall. In other words, the sound transducer element forms the outer surface of the solid structure and the sound field falls on at least a portion of the transducer element. The solid material may be a material to which the vibration sensor may be mechanically fixed, as compared to a medium such as air or water.

[0012] The system also includes at least one vibration sensor mounted to the sound transducing element, wherein the vibration sensor detects vibrations induced into the sound transducing element by the incoming sound field.

[0013] The system also includes at least one computing device configured to obtain sensor data from at least one vibration sensor, process the sensor data to generate output data, and provide the output data. In this regard, the computing device can be at least partially or completely implemented as a remote computing device and / or cloud computing resource.

[0014] The system may be configured to perform any method or combination of methods according to the present disclosure.

[0015] A computing device is provided, the computing device comprising at least one processor and a memory, the memory comprising instructions executable by the processor, wherein when the instructions in the processor are executed, the computing device is configured to perform the steps of any method or method combination according to the present disclosure.

[0016] A computer program or a computer program product and a computer-readable storage medium including program code are provided. The program code can be executed by at least one processor. When executing the program code, at least one processor performs any method or combination of methods according to the present disclosure.

[0017] For the system, the computing device and the computer program, advantages corresponding to those described for the method may be achieved.

[0018] It should be understood that the above features and the features to be explained below can be used not only in the indicated corresponding combinations, but also in other combinations or alone, without departing from the scope of the present disclosure. In particular, the features of the disclosed embodiments can be combined with each other in other embodiments.

[0019] It is to be understood that the features mentioned above and those yet to be explained below can be used not only in the respective combination indicated but also in other combinations or alone, without departing from the scope of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] These and other objects of the present invention will be appreciated and understood by those skilled in the art from the detailed description of the preferred embodiments and the following drawings, in which like reference numerals refer to like elements.

[0021] Figure 1 A vehicle including a plurality of vibration sensors mounted to an exterior vehicle panel is shown according to various examples.

[0022] Figure 2 The steps of a method for detecting an acoustic field incident on a solid structure according to various examples are schematically illustrated.

[0023] Figures 3 to 6 A vibration sensor mounted to a sound transducing element included in a solid structure is shown according to various examples. DETAILED DESCRIPTION

[0024] Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. It should be understood that the following description of the embodiments should not be understood in a limiting sense. The scope of the present invention is not intended to be limited by the embodiments or drawings described below, which are regarded as illustrative examples of general inventive concepts. The features of the various embodiments may be combined with each other unless otherwise specifically noted.

[0025] The drawings are to be regarded as schematic representations, and the elements shown in the drawings are not necessarily shown to scale. Instead, the various elements are represented so that their functions and general purposes become apparent to those skilled in the art. Any connection or coupling between functional blocks, devices, elements or other physical or functional units shown in the drawings and described hereinafter may also be achieved by indirect connections or couplings.

[0026] Hereinafter, techniques will be described that relate to detecting an acoustic field incident upon a solid structure, and in particular, detecting acoustic events and identifying the types of acoustic events.

[0027] Some examples of the present disclosure generally provide multiple sensors or other electrical processing devices. All references to circuits and other electrical devices and the functions provided by each are not intended to be limited to the contents shown and described herein. It should be recognized that any sensor or other processing device disclosed herein may include any number of microcontrollers, general-purpose central processing units (CPUs), graphics processor units (GPUs), integrated circuits, memory devices (e.g., flash memory, random access memory (RAM), read-only memory (ROM), electrically programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM) or other suitable variants thereof), and software that cooperates with each other to perform the operations disclosed herein. In addition, any one or more electrical devices may be configured to execute a program code embodied in a non-transient computer-readable medium programmed to perform any number of functions disclosed. In various instances, the processing device may be embodied as a remote or cloud computing device. It should be understood that other sensors may be used to detect vibrations in a solid-state body, including a sensor arrangement having an optical, mechanical, electromagnetic or capacitive structure, which may be used to detect vibrations in a solid-state body of a sound transducer element.

[0028] With the growth of safety, driver assistance technology and autonomous vehicles, the demand for acoustic and non-acoustic situations and event monitoring around the vehicle has grown widely. Due to the development of these technologies, it has become necessary to detect key sounds from outside the vehicle. Such sounds may include alarms, wake-up words or voice command words from emergency vehicles (e.g., police cars, fire trucks, ambulances), knocks from pedestrians or foreign objects hitting the car body, and other warning acoustic events occurring around the car (e.g., tire explosions, traffic accidents, shootings). For this reason, a given number of sensors need to be integrated in the vehicle to monitor and detect these acoustic events. These sensors must withstand the use of the car in all driving scenarios and are designed to withstand external factors (such as water, ice, dust or debris) that may change the performance of any sensor and endanger the performance of the intended system.

[0029] Compared to the most commonly used acoustic microphones that rely on a direct air path to sense sound, surface vibration sensors are typically sealed to the environment and can therefore be used in harsher environments. In addition, surface vibration sensors can be mounted on the B-side (or A-side) surface of vehicle components such as windows, car trim, doors, or other components. With this integrated solution, the sensor is not directly exposed to external agents such as water or ice. Therefore, surface vibration sensors can provide more reliable and consistent performance for automotive exterior sound event detection.

[0030] The present disclosure describes a method for detecting sound or acoustic events outside a car using a surface-mountable vibration sensor. Sound causes vibrations when it propagates through a medium. The sound-induced vibration motion can be sensed by an acoustic microphone (when the medium is air) or a vibration sensor such as an accelerometer or a dynamic strain / stress sensor (when the medium is a solid structure such as glass, plastic or metal surface or a composite material of a vehicle). The sensor output can be processed, analyzed, and compared with a sound feature library corresponding to selected key sound types. Finally, one or more specific types of sound events occurring outside the vehicle can be identified, and then corresponding warning decisions can be sent to the vehicle infotainment system (e.g., head unit, audio system) to remind or warn the driver and passengers.

[0031] Figure 1 A vehicle 100 is shown including a plurality of vibration sensors 102a - n mounted to an exterior vehicle panel 101 according to various examples.

[0032] like Figure 1 As can be seen in FIG. 1 , the vehicle 100 may be an example of a solid structure that is composed of a plurality of solid elements that together form a connection with the external environment (e.g., Figure 1The outer surface of the vehicle 100 is formed by a plurality of vehicle panels 101, which are body parts of solid materials, and on which vibration sensors 102a-n are mounted. The vehicle panel 101 may represent a sound transducing element of the vehicle 100. The outer surface of the vehicle panel 100 is a portion of the outer surface of the vehicle 100 that is visible and in direct contact with the surrounding environment. Therefore, an incoming sound field from the surrounding medium will generate an oscillating movement in the vehicle panel 101 because sound energy is introduced into these panels and propagates within the panels. In this example, the vibration sensor is mounted to the vehicle panel on the surface of the interior of the vehicle, so that the vibration sensor 102a-n is protected by the vehicle panel 101 from the external environment. The incoming sound field 103 is generated by a sound event outside the vehicle in the surrounding environment at a position in front of the vehicle 100. The incoming sound 103 propagates toward the vehicle 100 through the air and hits the outer surface of the vehicle. Vibration is generated at least in the vehicle panel 101 to which the vibration sensor is mounted by the incoming sound field. In other words, the incoming sound is induced into the sound transducing element of the solid structure, where it propagates as vibrations within the sound transducing element to the vibration sensors 102a-n. The vibration sensors 102a-n sense and detect the oscillatory movement (i.e., vibration) of the sound transducing element and output measurement data representative of the vibration.

[0033] The vehicle 100 may also include a computing device configured to process the vibration sensor data. The vehicle 100 may also include a network interface that may be used to transmit the sensor data to a remote processing device or a cloud-based processing device via a communication network.

[0034] Figure 2 The steps of a method for detecting an acoustic field incident on a solid structure according to various examples are schematically illustrated.

[0035] The method starts at step S10.

[0036] In step S20 , sensor data are obtained from at least one vibration sensor mounted to the solid structure, wherein the sensor data correspond to measurement data of vibrations generated in the solid structure, in particular in a sound transducing element of the solid structure, by an incoming sound field.

[0037] In various examples, external sound events can be monitored by one or more vibration sensors 102 mounted on the surface of the vehicle 100, i.e., sound transducing or transmitting vehicle components having an exterior surface 101, such as a windshield, metal body panels and roof, and plastic bumper panels. The vibration sensor 102 can be in the form of an accelerometer or a dynamic strain / stress sensor that measures acceleration, deformation, and / or force caused by structural vibration. In the case of an accelerometer, the sensing axis is preferably oriented perpendicular to the mounting surface, while in the case of a dynamic strain / stress sensor, the sensing direction is parallel to the mounting surface (i.e., in a plane). In other words, the vibration sensor mounted on the vehicle surface picks up the sound signal in the sound transducing element and outputs sensor data, i.e., a sensor measurement signal, which represents the vibration detected in the sound transducing element 101.

[0038] In step S30 , the sensor data is processed to generate output data.

[0039] In various examples, the sensor output signal is fed into an acoustic signal processing and feature extraction module (ASPFEM). The ASPFEM analyzes the signal, extracts features (such as time domain and frequency domain characteristics of the signal), compares the extracted features to a library of sound features corresponding to selected key sound types (which is predetermined knowledge), and identifies sound events. The identified sound events (e.g., sound event types) can be the output of the processing of the sensor data.

[0040] In step S40 , the determined output data is provided. For example, the output data may be provided to a central processing device of the vehicle, wherein the output data may be further used, for example, for safety applications.

[0041] In optional step S50, a warning message corresponding to the determined sound event type is provided to the user. For example, the system outputs a warning decision to the vehicle infotainment system to warn the driver and passengers inside.

[0042] The method ends in step S60.

[0043] In addition to detecting the type of emergency sound event outside the car, the relative incoming direction of the sound event can also be determined by comparing multiple channels of signals from surface vibration sensors installed at various locations of the vehicle (such as the front, rear, left and right sides of the vehicle).

[0044] The sensor data may be provided directly to a processing device, which may be, for example, a computing device or processor, or may be temporarily stored in a memory and provided from the memory to the computing device. The sensor data may also be transmitted to a remote and / or cloud computing resource for processing, where the output data may be transmitted back to the vehicle via a wired or wireless communication network.

[0045] Processing may include determining sound events and / or types of sound events that generate the sound field.

[0046] The sounds or acoustic events picked up and detected by this method may correspond to one or more of the following sound event types:

[0047] Emergency vehicle detection, where information and / or warnings are provided based on the presence of emergency vehicle sounds, sirens, or other acoustic warnings in the surrounding environment, and / or what type of first responders are present.

[0048] Events happening around the vehicle, such as pedestrians around the car, workers or first responders, accidents, or ongoing construction around the road.

[0049] Communication events, such as human-to-machine communications (including wake-up words or more complex commands) and / or inside-outside communications (including, for example, conversations between the outside and inside of a car or building).

[0050] In various examples, processing can include processing the sensor data in a neural network to directly generate output data. In various examples, applying the training function can include applying an end-to-end training network.

[0051] Figures 3 to 6 A vibration sensor 102 is shown mounted to a sound transducing element 101 according to various examples.

[0052] like Figure 3 As can be seen in FIG. 1 , the sound transducing element 101 is depicted without the solid structure 100 of which it is a part. For example, the sound transducing element 101 may be secured to other parts of the solid structure 100. In various examples, the sound transducing element 101 may be a vehicle panel.

[0053] The front side of the sound transducing element 101 may be exposed to the external environment, including a gas or fluid medium, wherein the sound field 103 propagates toward the front side of the sound transducing element 101. When the incoming sound field 103 hits the front surface of the sound transducing element 101, vibrations are induced in the sample using the element 101.

[0054] The vibration sensor 102 is mounted on the back side of the sound transducer element 11. The vibration sensor 102 may be mounted on the surface opposite the outer surface impacted by the incoming sound field 103. In this regard, the sound transducer element 101 may have a certain wall thickness that separates the vibration sensor from the external environment. As depicted by the arrows, the vibration sensor may be fixed, in other words, mechanically connected to the mounting surface of the sound transducer element 101 at various locations along the mounting surface, which will result in different amplitudes at different frequencies detected by the vibration sensor, depending on the material of the sound transducer element used.

[0055] exist Figure 3 In an example, an acceleration sensor may be used that senses vibrations of the sound transducing element 101 in a sensing direction perpendicular to the mounting surface and the impact surface of the incoming acoustic field impact. In other examples, a strain sensor may be used that can measure at least two-dimensional strain along a mounting surface of the sound transducing element 101 to which the strain sensor may be fixedly mounted.

[0056] Figures 4 to 6 The elements in correspond to Figure 3 The components are described above and therefore will not be described in detail. Figure 4 As can be seen in FIG. 1 , the vibration sensor 102 can also be mounted on the side surface of the sound transducer element 101. Figure 5 As can be seen in FIG. 1 , the vibration sensor 102 can also be mounted on an area of ​​the sound transducer element 101 which is separated from the sound entry surface by a mounting point to which another vehicle component 104 is fixed. In this regard, for example, the vibration sensor 102 can be mounted on an elongated arm connected to the sound transducer element 101. This may have the advantage that certain frequencies can be amplified. Figure 6 As can be seen in FIG. 1 , the vibration sensor 102 can be mounted to an area of ​​the sound transducer element 101 that is inside a solid structure, such as a free space inside a solid structure, which is not in contact with the external environment. The vibration sensor can provide more design flexibility so that it can be arranged in a location that is more easily sealed and protected from external influences such as physical contact, humidity, corrosion, and wind.

[0057] Based on the above discussion, some general conclusions can be drawn:

[0058] The solid structure may be composed of a plurality of elements formed of solid material (in other words, solid-state material), which may be fixed to each other to form the solid structure. In various examples, such a solid structure may form an outer surface that may be directly exposed to an external environment, i.e., a sound transmission medium, which is not a solid medium, but may be a gas or a liquid, in which a sound event may occur. Such a sound event may generate an incoming sound in the medium (usually air or water), which may propagate through the gas or liquid medium to impinge on the solid structure.

[0059] The solid structure may include a sound transducer element, which may be one of the solid elements constituting the solid structure. The sound transducer element may be made of a solid material. The sound transducer element may have an outer surface, which may be part of the outer surface of the solid structure. In this regard, the sound field falling on the outer surface of the solid structure also falls at least partially on the outer surface of the sound transducer element. The outer surface of the sound transducer element may be a flat surface, such as a planar surface, or a curved surface, such as a free-form surface. The sound field may directly hit the sound transducer element, in other words, there will be no other mechanical elements, such as a grid or a grille, between the sound event location and the sound transducer element. The sound transducer element may be a monolithic element. The sound transducer element may have a wall thickness greater than 0.1 mm, or greater than 0.5 mm, or greater than 1 mm. The sound transducer element may be made of glass, metal, plastic or a composite material. The sound transducer element may have the form of a sheet or plate, and may, for example, exceed an area of ​​20 mm×20 mm in at least one or two directions (along the front surface or the rear surface, i.e. in a plane substantially perpendicular to the surface normal).

[0060] The sound transducing element may be, for example, a panel of a vehicle, such as the outer surface of a door, a glass window of a vehicle, an outer panel of a building, a front trim of an electronic display device, and similar sheets or plates that form a user-accessible, and / or touchable, and / or visible outer surface of a solid structure. In other words, such a sound transducing element may have an outermost outer surface facing the external environment (e.g., facing the user), wherein no other portion of the solid structure is arranged between the sound event location and the sound transducing element.

[0061] When the sound field falls on the sound transducer element, when the sound energy is introduced into the sound transducer element, vibrations can be generated in the sound transducer element and propagated within the sound transducer element, which is caused by the incoming external sound field. In this regard, the surface vibration of the mounting surface on which the vibration sensor is mounted can represent the incoming sound that impacts the sound transducer element. The sound field may not directly hit the vibration sensor, but the vibration sensor can be arranged in a position where the incoming sound cannot be directly propagated. In other words, the incoming sound can propagate in the air toward the solid structure and can hit the outer surface of the sound transducer element, whereby the incoming sound field can induce vibrations in the sound transducer element, and the vibrations can propagate from the outer surface through the sound transducer element to the mounting surface of the sound transducer element, and the vibration sensor can be fixed to the mounting surface, wherein the vibration sensor detects the oscillatory movement of at least a portion of the mounting surface. In this regard, the outer surface of the solid structure and the sound sensing element can be a surface facing the direction in which the sound event occurs and / or away from the center or center of gravity of the solid structure.

[0062] In various examples, since the vibration sensor can be in direct mechanical contact with the mounting surface, the vibration propagating in the sound transducer element can be directly transmitted from the mounting surface to the vibration sensor. The outer surface can correspond to a flat surface directly facing the surrounding environment (particularly the surrounding air). In various examples, the outer surface of the sound transducer element can correspond to a display surface on which information or warnings can be displayed. In various examples, the outer surface of the sound transducer element can correspond to the surface of a solid structure that can be touched by a user (such as a touch input) or a surface that is visible to a user. The area of ​​the outer surface of the sound transducer element on which the sound field falls to generate vibrations can be greater than 1 cm 2 , greater than 5cm 2 , greater than 20cm 2 , greater than 100cm 2 or more than 1000cm 2 The outer surface of the sound transducing element may have a concave curvature, or may have a convex curvature, or may be a planar surface.

[0063] In various examples, the complete vibration sensor can move in oscillatory motion (i.e., vibration) with the sound transducing element (particularly the mounting surface to which the sensor is mounted). For example, the vibration sensor can be an accelerometer that is mechanically fixed to the transducing element and thus moves with the mounting surface of the sound transducing element in an oscillatory motion induced into the sound transducing element by the incoming sound field.

[0064] In various examples, an incoming sound field impinging on the sound transducer element may represent a mechanical stress on the sound transducer element, wherein vibrations in the sound transducer element induced by the incoming sound field may correspond to mechanical strains in the sound transducer element, which may be measured by a vibration sensor. The vibration sensor may be a strain sensor that detects oscillating material strains induced into the sound transducer element by the incoming sound field. At least a portion of a surface of the vibration sensor may be secured to a mounting surface of the sound transducer element such that the surface of the vibration sensor is in tension with the sound transducer element.

[0065] The method also includes obtaining sensor data from at least one vibration sensor mounted to a surface of the sound transducing element, wherein the vibration sensor detects vibrations of the mounting surface of the sound transducing element, the vibrations being induced into the sound transducing element by the incoming sound field.

[0066] The method may also include processing the sensor data to generate output data, and providing the output data.

[0067] In various instances, processing sensor data may include extracting a feature set from the sensor data, wherein the sensor data represents a sound event that generates a sound field; determining a type of the sound event by comparing the extracted feature set to a plurality of reference feature sets that each represent a different type of sound event; and providing the sound event type of the sound event that generates the sound field as output data.

[0068] In various examples, processing the sensor data may include determining a warning message for the type of acoustic event and outputting the warning message to a user of the solid structure.

[0069] In various examples, processing the sensor data may include reconstructing a sound field based on the sensor data, and outputting (ie, replicating) the reconstructed sound field so that a user can hear the replicated sound field as if in the original sound field.

[0070] In various examples, processing the sensor data may include identifying physical contact of a solid structure and / or a sound transducer element with another solid body and / or a gas and / or a liquid, such as a touch and / or an impact and / or a collision into the solid structure and / or the sound transducer element. Such physical contact may be detected, identified and / or classified, wherein corresponding contact information may be provided as output data, such as including classification information or a contact type of the physical contact.

[0071] In various examples, processing the sensor data can include distinguishing between physical contact and an acoustic field as described above, both of which can generate vibrations detected by a vibration sensor in a sound transducer element. Information about the acoustic field can be provided separately from information about the physical contact.

[0072] In various examples, processing the sensor data may include applying a training function to the sensor data, wherein the training function is trained using a training sensor data set and corresponding known reference output data. The training function may, for example, include a neural network, wherein the sensor data is fed into the neural network to directly generate output data. In various examples, applying the training function may include applying an end-to-end training network.

[0073] For example, the training function can be an end-to-end training function that is trained using multiple training data sets. The training data set can include input data associated with reference output data. Applying the training function can be performed by a neural network that can include multiple classifier functions. In various instances, the training function can include one or more of known machine learning classifiers. Without limitation, the training function can be based on one or more of, for example, a support vector machine, a decision tree and / or a Bayesian network, k-means clustering, Q learning, a genetic algorithm and / or an association rule. For example, the neural network can be a deep neural network, a convolutional neural network, or a convolutional deep neural network, an adversarial network, a deep adversarial network and / or a generative adversarial network, or a model-based machine learning network architecture.

[0074] The output data may include the direction in which the sound event from which the sound field originates is located.

[0075] The vibration sensor may be mounted to a surface opposite the outer surface onto which the acoustic field falls.

[0076] The vibration sensor may be sealed by the sound transducing element from the external environment, including the atmosphere from which the sound field originates and which is transmitted into the medium in which the sound propagates.

[0077] A plurality of vibration sensors may be mounted to the solid structure, to a corresponding plurality of sound transducing elements, and wherein the orientation of the surface normals of two sound transducing elements receiving the sound field spans at least 90°, preferably 180°.

[0078] According to the principles of the present disclosure, the larger the external impact surface of the sound transducer element, and the thinner the wall thickness, for example in the direction of the incoming sound field, the clearer the measurement signal generated by one or more vibration sensors attached to the solid body may be. The measurement signal (i.e., sensor data) can be processed to identify the type of sound event, or even to reconstruct the sound field to be broadcast, for example, in the interior of a car. Processing the sensor data can, for example, include performing a speech / voice recognition method based on the sensor data.

[0079] In various examples, an acoustic field as a space can be generated and originate from acoustic events that occur in a medium and generate sound. Sound of the acoustic field can propagate in the medium, and specific incoming sound waves can have directions in which they propagate. In various examples, incoming sound can strike a solid surface from any direction and be picked up by an acoustic transducer element, for example in the form of vibrations in the acoustic transducer element.

[0080] The sound transducer element may be essentially a 3D solid structure, wherein one dimension is significantly smaller than the dimensions in the other two directions, which may be referred to as the thickness of the sound transducer element. In other words, the outer surface that will be impacted by the incoming sound field may extend substantially along a transverse plane (or in the x-direction and the y-direction), which may have a surface normal pointing to the exterior of the solid structure, wherein the thickness may extend in the direction of the surface normal or in the z-direction.

[0081] A sound event may correspond to a sound source, which in some instances may be considered a point source in nature. If a line is drawn from the point source (i.e., the sound event) to the surface location where the vibration sensor is mounted, the maximum signal strength may be obtained when the line is exactly in the direction of the normal direction (i.e., the thickness direction) of the surface.

[0082] Advantages of the disclosed technology include providing a solution for detecting emergency sound events outside a car using a vibration sensor. Classic sound sensing microphones are unlikely to withstand the harsh environmental conditions outside a car, making traditional acoustic microphones difficult to be a practical solution for such applications. Vibration sensors can be environmentally sealed and are not affected by water, snow, ice, dust, etc.

[0083] Furthermore, the vibration sensor works by sensing sound-induced structural vibrations. It can be mounted at any convenient location on the vehicle surface, either on the exterior (A side) or interior (B side), making it more flexible to integrate into the vehicle than integrating an acoustic microphone into the vehicle.

[0084] Another benefit that has been demonstrated by preliminary experimental data is that surface vibration sensors are less susceptible to wind noise than traditional acoustic microphones. This is an important advantage for exterior sound detection applications when the car is moving at moderate to high speeds.

[0085] In various examples, processing sensor data may include detecting and distinguishing physical contact between an outer surface of a solid structure and / or a sound transducer element and another solid body or fluid (e.g., rain) or wind, wherein events including touches, impacts, or deformations of the solid structure and / or the sound transducer element may be more easily and clearly distinguished from pure sound fields. Thus, for example, touches to a surface, or impacts or collisions to a solid structure and / or a sound transducer element (which result in sound and / or vibrations), and pure sound events may be more easily detected and identified than with conventional microphones. In various examples, such a vibration sensor may be used together with a conventional microphone, and the signals provided by both may be processed in parallel to more reliably detect sound fields and sound events.

[0086] In a preferred example, the sound field around the vehicle can be detected, wherein the sound transducer element can be a panel or decorative part of the vehicle, such as a panel or decorative part of the vehicle body, which forms the outer surface of the vehicle. In contrast to, for example, a grid, the outer surface can be a continuous surface. The outer surface can be a surface affected by a sound transmission medium (such as air), and a sound event can occur in the sound transmission medium. In various examples, the interior of a vehicle or building can also be considered as a space where a sound event may occur, so that the decorative part of the vehicle or building facing the sound event in the interior space or room can form an impact surface (i.e., corresponding to the outer surface) to be exposed and impact the sound field. In various examples, the outer surface of the sound transducer element can be a surface in contact with the surrounding gas atmosphere, for example, with a direct sound path to the outer surface, wherein the sound falls directly on the external sound impact surface. In this regard, vibrations can be induced in the sound transducer element at the interface between the surrounding gas or liquid medium (such as air) and the sound transducer element. In various examples, the sound transducer element can also be a decorative part inside the vehicle, wherein the sound field can originate from the outside or inside of the car, such as a conversation or command of a passenger.

[0087] Although the disclosed technology has been described with respect to certain preferred embodiments, equivalents and modifications will occur to those skilled in the art upon reading and understanding this specification. The present disclosure includes all such equivalents and modifications and is limited only by the scope of the appended claims.

[0088] For illustration, various scenarios are disclosed above in connection with vehicles. Similar techniques can be readily applied to other classes and types of solid systems, such as, for example, buildings, electronic consumer devices, or any kind of outdoor or indoor structure that may include surfaces of solid materials that are exposed to and receive acoustic energy from an external sound field.

Claims

1. A method for detecting an acoustic field (103) around a solid structure (100), wherein the solid structure (100) comprises at least one acoustic transducer element (101), the at least one acoustic transducer element being part of the solid structure (100) and forming at least a part of an outer surface of the solid structure (100), wherein the acoustic field (103) at least partially falls on the acoustic transducer element (101), the method comprising: include: obtaining sensor data from at least one vibration sensor (102) mounted to a mounting surface of the sound transducer element (101), wherein the vibration sensor (102) detects vibrations in the sound transducer element (101), the vibrations being induced into the sound transducer element (101) by an incoming sound field (103); processing the sensor data to generate output data; and The output data is provided.

2. The method of claim 1, wherein the vibration sensor (102) is an accelerometer that moves with the mounting surface of the sound transducer element (101) in an oscillatory movement induced into the sound transducer element (101) by the incoming sound field (103).

3. The method of claim 1, wherein the vibration sensor (102) is a strain sensor that detects oscillating material strain induced into the sound transducing element (101) by the incoming sound field (103).

4. The method according to claim 1, wherein the solid structure (100) is a vehicle, a building or an electronic consumer device, and the outer surface of the sound transducer element (101) is a flat continuous surface, and the incoming sound field (103) impinges on the flat continuous surface.

5. The method of claim 1, wherein processing the sensor data include: extracting a set of features from the sensor data, wherein the sensor data represents sound events that generated the incoming sound field (103), determining the type of the sound event by comparing the extracted feature set with a plurality of reference feature sets each representing a different sound event type, and The type of the sound event is provided as output data.

6. The method of claim 5, wherein processing the sensor data further comprises: include: determining a warning message of said type of said sound event, and The warning information is output to a user of the solid structure (100).

7. The method of claim 1, wherein processing the sensor data include: A training function is applied to the sensor data, wherein the training function is trained using a training sensor data set and corresponding known reference output data.

8. The method of claim 1, wherein the output data comprises a direction in which a sound event from which the sound field (103) originates is located.

9. The method according to claim 1, wherein the vibration sensor (102) is mounted to a surface of the sound transducer element (101) opposite to the outer surface on which the sound field (103) strikes, or is mounted on the outer surface of the sound transducer element (101) on which the sound field (103) strikes.

10. The method according to claim 1, wherein the sound energy of the incoming sound field (103) propagates in the air toward the solid structure (100) and falls on the outer surface of the sound transducer element (101), whereby the incoming sound field (103) induces vibrations in the sound transducer element (101), and the vibrations propagate from the outer surface through the sound transducer element (101) to the mounting surface of the sound transducer element (101), and the vibration sensor (102) is fixed to the mounting surface, wherein the vibration sensor (102) detects the oscillatory movement of at least a portion of the mounting surface.

11. The method according to claim 1, wherein the sound transducing element (101) has / is formed as a sheet made of glass, plastic, metal or a composite material and exceeds an area of ​​20 mm x 20 mm.

12. The method of claim 1, wherein the vibration sensor (102) is at least partially isolated by the sound transducing element (101) from an external environment from which the sound field (103) originates.

13. A method according to claim 1, wherein a plurality of vibration sensors (102a-n) are mounted to a corresponding plurality of sound transducer elements (101) in the solid structure (100), and wherein the orientation of the surface normal of the sound transducer element (101) receiving the incoming sound field (103) spans at least 90°, preferably 180°.

14. A system (103) for detecting a sound field, include: A solid structure (100) comprising at least one sound transducer element (101) formed of a solid material, wherein the sound transducer element (101) forms at least a portion of an outer surface of the solid structure (100), and the sound field (103) impinges on at least a portion of the at least one sound transducer element (101); at least one vibration sensor (102) mounted to the sound transducer element (101), wherein the vibration sensor (102) detects vibrations induced into the sound transducer element (101) by an incoming sound field (103); as well as At least one computing device is configured to obtain sensor data from the at least one vibration sensor (102), process the sensor data to generate output data, and provide the output data.

15. The system according to claim 14, further configured to perform the method according to any one of claims 2 to 13.