Method and vehicle for detecting organisms in a vehicle compartment

Abstract

The present invention relates to a method for detecting a living organism (1) in a vehicle interior (2), characterized by the following method steps: a) implementing a NAF database (5) in a computing unit (4), the NAF database (5) including a number of NAFs of the vehicle interior (2) for different presence scenarios, b) recording background noise in the vehicle interior (2) using a microphone (3) at least when a vehicle door is closed, c) comparing the recorded background noise with the NAFs stored in the NAF database (5) by the computing unit (4), d) calculating at least one NAF that has the highest consistency with the recorded background noise, and e) determining the presence of a living organism (1) in the vehicle interior (2) by the computing unit (4) for each presence scenario based on the NAFs calculated in step d).

Description

[Technical Field]

[0001] The present invention relates to a method for detecting living organisms in a vehicle compartment of the kind defined in detail in the preamble of claim 1, as well as to a vehicle of the kind defined in detail in the preamble of claim 10. [Background technology]

[0002] In everyday life, children and animals are sometimes left behind or intentionally left unattended in vehicle cabins. For example, a dog owner may not take their dog with them when they want to go out for a quick grocery run. Especially in strong summer sun or cold winter temperatures, the interior of a vehicle can become very hot or very cold, which can cause children or pets to overheat or suffer from hypothermia, potentially creating a life-threatening situation. Therefore, a means and method for alerting a vehicle user that a child or animal has been left behind in the vehicle cabin is needed.

[0003] For example, German Patent Application Publication No. 102019212412 discloses a monitoring device and method for monitoring living creatures in a vehicle cabin. This document describes monitoring the vehicle cabin using visual detection means such as a camera, acoustic detection means such as a microphone, and a temperature sensor. Data generated by the mentioned camera, microphone, and temperature sensor are evaluated by a computing unit using artificial intelligence. The computing unit performs classification to identify the respective presence status. Data from additional sensors, such as an occupancy sensor, a motion sensor, and / or a seat belt sensor, can also be taken into account to determine whether a living creature is present in the vehicle cabin. Using the camera and microphone, as well as the presence of a display and speakers in the vehicle, a baby monitor function can be provided. Thus, video and audio content recorded in the vehicle cabin can be transmitted to a vehicle user's smartphone, and audio and video content recorded on the smartphone can be transmitted to the vehicle. This allows the vehicle user to reassure their child in the vehicle cabin. However, this monitoring device is relatively complex and expensive due to the need to provide various types of sensors.

[0004] It is therefore desirable to provide the ability to detect the presence of living organisms within a vehicle cabin using a simple, low-cost means.

[0005] The detection of distance information by sound is well known, and for this purpose there exist various devices such as echo sounders, sonar or ultrasonic sensor systems, which are used to form parking assistance systems, also commonly called "parking buzzers".

[0006] Additionally, spatial modeling of ambient acoustic characteristics is also a current research topic. Andrew Luo et al. describe methods and means for obtaining structural information about a scene purely through acoustics. To this end, the researchers present a concept called the neural acoustic field (NAF). Such a neural acoustic field contains geometric information that describes the structure of the measured surroundings. To generate such a neural acoustic field, acoustic measurements are performed in the surroundings to be mapped, such as a room or a residence. This acoustic measurement requires a large number of individual measurements. In each individual measurement, sound sources and acoustic receivers are arbitrarily positioned around the object under test, and an acoustic pulse is emitted via the sound source, which is detected by the acoustic receiver. The measured pulse response is converted into a spectrogram using a short-time Fourier transform. The spectrogram thus generated is then processed by an artificial neural network to determine the examined neural acoustic field as an output value. For more information, see: Learning Neural Acoustic Fields, Andrew Luo, Yilun Du, Michael J. Tarr, Joshua B. Tenenbaum, Antonio Torralba, Chuang Gan, https: / / doi.org / 10.48550 / arXiv.2204.00628. Summary of the Invention [Problem to be solved by the invention]

[0007] The present invention is based on the object of providing an improved method for detecting living organisms in a vehicle interior, which works more reliably with simpler hardware.

[0008] According to the invention, this object is achieved by a method for detecting living organisms in a vehicle compartment having the features of claim 1. Advantageous embodiments and developments, as well as a vehicle for implementing the method, emerge from the claims dependent thereon.

[0009] In a standard method for detecting living organisms in a vehicle cabin, the vehicle cabin is acoustically monitored by at least two microphones, and the signals generated by the microphones are evaluated by a computing unit using artificial intelligence. According to the present invention, this method comprises: a) implementing a NAF database in a computing unit, the NAF database comprising a number of neural acoustic fields in the vehicle cabin for various occupancy scenarios, the NAF database being generated by an initial measurement campaign in a representative vehicle cabin, recording at least one neural acoustic field in each case, varying at least the type of dummy, the number of dummies, and the seat position of the dummies; b) recording background noise in the vehicle cabin with a microphone at least when the vehicle door is closed; c) comparing the recorded background noise with the neural acoustic fields stored in the NAF database by a computing unit; d) determining by a calculation unit at least one neural sound field that has the highest match to the recorded background noise; e) determining, by the computing unit, the presence of a living being in the vehicle cabin in response to each presence scenario based on the neural acoustic field calculated in step d). It has been further developed by

[0010] The method according to the present invention allows acoustic monitoring of the vehicle interior in a completely new way, which allows the presence of living things in the vehicle interior to be detected more reliably using simple means. The neural acoustic field represents an acoustic mapping of the vehicle interior. To detect living things in the vehicle interior, corresponding maps are generated for all possible presence scenarios. These measurement campaigns are carried out, for example, by the vehicle manufacturer.

[0011] To simulate the presence of living beings, dummies such as crash dummies, dolls, or containers filled with water are used. Adults, children, and animals can be abstracted by various dummy types, e.g., large or small dummies. The dummies can also be covered with blankets, feathers, or hair to simulate fur or feathers. The number of dummies can vary between zero and the maximum number of seats in the vehicle interior. Dummies are then assigned to all possible seating positions. This ensures that all possible presence scenarios are covered by the neural acoustic field.

[0012] When the vehicle is subsequently used, for example with an adult driver in the driver's seat and a child and a dog in the rear seat, the corresponding acoustics are generated within the vehicle cabin. These acoustics are then mapped by the corresponding neural acoustic fields. In this case, they are correlated with the neural acoustic fields from the initial measurement campaign, with a presence scenario in which a large dummy is placed in the driver's seat and two small and two specially small dummies are placed in the left and right rear seats of the vehicle. In the NAF database, information on the type of dummy, the number of dummy occupants, and the seating positions of the dummy occupants on which each presence scenario is based is appropriately linked to the respective neural acoustic fields.

[0013] To ensure the highest possible correspondence between the measured background noise and the neural acoustic fields stored in the NAF database, the corresponding measurement campaigns are carried out individually on a variety of representative vehicles. "Representative" in this case means that the vehicles used to carry out the initial measurement campaign correspond to at least the same model series as the vehicle to which the method will ultimately be applied. This ensures that the cabin geometry, which is important for the cabin acoustics, is substantially the same.

[0014] The initial measurement campaign can be designed to be as comprehensive as desired, for example, further influencing factors can be investigated, such as the seat adjustment, i.e. in particular the angle of inclination of the backrest, the overall position or displacement of the vehicle seat, the presence or absence of a child seat and / or further objects in the vehicle. In each case, in addition to the dummy type, number of dummy and seat position of the dummy, the neural sound fields corresponding to different seat adjustments, e.g. a handbag placed at the rear of the vehicle or a laptop bag placed on the passenger seat, can be recorded.

[0015] At least one nerve sound field is recorded for each parameter combination in the initial measurement campaign. However, preferably, multiple nerve sound fields may be recorded, thereby reducing measurement inaccuracies or errors resulting from statistical events. In particular, multiple nerve sound fields are recorded for different source configurations for the same presence scenario. For this purpose, in particular, the type of sound source and / or its position relative to the vehicle cabin are changed.

[0016] The use of the sound of a vehicle door closing as a sound source to detect background noise within the vehicle cabin is particularly practical for later use. Since a short, distinctive acoustic pulse is generated when a vehicle door closes, the sound of a vehicle door closing is particularly suitable as a sound source. Furthermore, there is no need to output an artificial stimulus signal within the vehicle cabin. In this context, "during closure" means that the relevant acoustic signal is analyzed during the time interval from when the vehicle door closes to immediately thereafter. This time interval may range from a few seconds to a few milliseconds. The microphone must be active for at least the duration of the recording period. The microphone can also be activated before the actual recording period and deactivated any time after the end of the recording period. For example, the microphone can be activated when the sensor determines that the vehicle door is open. When a vehicle door is opened or closed, the vehicle is usually stationary. This is the moment when a vehicle occupant enters or leaves the vehicle. This is a critical moment for identifying the presence of a living being within the vehicle cabin. However, it is generally also possible to activate the microphone and record audio signals while the vehicle is moving, as will be discussed below.

[0017] By providing multiple microphones in the vehicle cabin, multiple neural sound fields can be recorded simultaneously for one acoustic signal source. The mounting positions of corresponding microphones are correspondingly consistent between a representative vehicle and a vehicle in which the method is implemented. Typically, a vehicle is equipped with multiple such cabin microphones, for example, to provide a hands-free device for phone calls or to detect voice commands for a voice assistant. In order to be able to detect the voices of different vehicle occupants as much as possible, the microphones are usually provided at various mounting positions.

[0018] The calculation unit calculates in step d) at least one neural acoustic field that best matches the recorded background noise. The comparison of the background noise described in step c) can in particular be provided in such a way that at least one neural acoustic field is generated from the recorded background noise and compared with the NAF database, such that such a field stands out in particular with respect to the total neural acoustic fields stored in the NAF database, thereby revealing what presence scenarios are present, i.e., how many living beings are present in the vehicle cabin, in which seat positions, what type of living beings they are, i.e., for example, adults, children, or animals.

[0019] By processing the acoustic signals, the computing unit can also identify the type of animal if desired, for example by distinguishing between barks, meows, chirps, etc.

[0020] However, it is possible that multiple neural acoustic fields in the NAF database have similarities to the recorded background noise. In this case, the computing unit compares information linked to each presence scenario. For example, if two living creatures are present in the vehicle cabin, the neural acoustic fields may show a high degree of agreement. However, if the respective seat positions are different, the method of the present invention can determine the number of living creatures present in the vehicle cabin, but not their exact seat positions.

[0021] However, the method according to the present invention at least makes it possible to distinguish whether or not one or more organisms are present in the vehicle cabin.

[0022] In an advantageous development of the method according to the invention, the sound of a vehicle door slamming is used as the acoustic signal source in the initial measurement campaign for generating the NAF database, and in particular the neural acoustic field is also recorded with variations of the slamming vehicle door. Since the method according to the invention for identifying the presence of a living being in a vehicle interior is carried out when the vehicle door is opened and closed, at least in a stationary state of the vehicle, the sound of a vehicle door slamming is used as the sound source. Therefore, the sound of a vehicle door slamming is also used in a particularly advantageous manner as the sound source for generating the NAF database in the initial measurement campaign. The neural acoustic field generated in the initial measurement campaign is therefore based on the same sound source as the sound source used when the method is subsequently carried out in the vehicle. This allows for a high degree of correspondence between the background noise recorded in the vehicle interior (i.e., in particular the neural acoustic field derived from the background noise) and the neural acoustic field stored in the NAF database.

[0023] According to a further advantageous embodiment of the method, the calculation unit determines which vehicle door has been closed, taking into account the door closing sensor. Depending on which vehicle door is closed and used as the sound source, the position of the sound source changes, which accordingly affects the corresponding neural sound field. Therefore, a separate neural sound field is recorded for each possible combination of slammed vehicle doors. Here, a single vehicle door may be slammed, or multiple vehicle doors may be slammed simultaneously or in close succession. This is taken into account appropriately in the initial measurement campaign. Taking into account the data transmitted from the door closing sensor, the calculation unit can accurately detect which vehicle door has been slammed. The calculation unit can then take this information into account when calculating a match between the background noise recorded in the vehicle cabin and the neural sound fields stored in the NAF database. Therefore, the calculation unit selects only neural sound fields from the NAF database that are suitable for a match when the corresponding vehicle door is used as the sound source.

[0024] Furthermore, in a further advantageous embodiment of the method according to the present invention, different initial measurement campaigns are performed to generate the NAF database depending on the vehicle equipment. In addition to the influencing factors already mentioned, the vehicle equipment also influences the respective neural sound field. For example, if a particular vehicle seat type, such as sports seats, is installed and these seats have a specific material composition, such as leather, or if other special components, such as burl wood trim, are installed in the vehicle interior, this can affect the sound dispersion in the vehicle interior. For example, smooth surfaces reflect sound differently than rough surfaces, such as velvet or textured plastic.

[0025] This means that a vehicle manufacturer can perform corresponding measurement campaigns not only for different model series but also for different vehicle equipment within each model series. All of the generated nerve acoustic fields can then be implemented in the vehicle's computing unit, or only nerve acoustic fields recorded in a representative vehicle interior with the corresponding vehicle equipment can be implemented in the vehicle's computing unit. This saves memory capacity on the computing unit. However, if the vehicle equipment changes over the vehicle's service life, missing NAF databases or nerve acoustic fields must be retrofitted into the computing unit. This can be done, for example, by updating, preferably via an over-the-air update.

[0026] According to a further advantageous embodiment of the present invention, the computing unit performs a measurement campaign to expand the NAF database during the vehicle's use phase, and information describing the current presence scenario is actively queried by the user. In other words, the computing unit is capable of learning. The vehicle user can record new neural acoustic fields during the use phase, i.e., during the vehicle life cycle phase during which the vehicle is used by the user. For this purpose, dummies are no longer used, but rather real living beings, such as family members or pets, who typically travel with the vehicle. This further increases the degree of correspondence between the background noise recordable in the vehicle and the neural acoustic fields stored in the NAF database. In order for the computing unit to know the corresponding presence scenario, this information must be manually provided to the computing unit. For this purpose, the user can use an appropriate human-machine interface, such as a touch-sensitive display.

[0027] In a further advantageous embodiment of the method according to the present invention, method steps b) to e) are performed again after the vehicle door is closed, with an acoustic stimulus signal being output from at least one speaker arranged in the vehicle. As already mentioned, the method according to the present invention can be used to detect the presence of living beings in the vehicle interior not only when the door is slammed, but also at a later time, for example, while the vehicle is moving. This makes it possible to verify the presence of previously detected living beings, in particular their number, type, and seat position. Since vehicle doors should not be opened or closed while the vehicle is moving, a separate sound source is preferably used. For this purpose, vehicle speakers can be used, in particular speakers arranged at various locations in the vehicle interior that output audio signals. Particularly preferably, the stimulus signal here corresponds to the stimulus signal used in the initial measurement campaign.

[0028] For example, when the vehicle is stopped, pets can enter the vehicle through an open door window, so the correct number of animals can be calculated even while the vehicle is moving.

[0029] Preferably, the acoustic stimulus signal output by the at least one speaker contains only frequencies in the ultrasonic and / or infrared frequency ranges. This has the advantage that the corresponding stimulus signal is not perceptible by the vehicle occupants. This makes it possible to carry out the method according to the invention without the vehicle occupants noticing. This particularly improves user comfort.

[0030] In a further advantageous embodiment of the method according to the invention, the computing unit also uses artificial intelligence in method step d) to identify a neural sound field that matches the background noise. For this purpose, proven methods and procedures from artificial intelligence, such as artificial neural networks, in particular deep artificial neural networks, can be used. This allows a more reliable determination of the match between the recorded background noise and the neural sound fields stored in the NAF database. In particular, a corresponding machine learning model compares the neural sound field derived from the recorded background noise with the NAF database.

[0031] According to a further advantageous embodiment of the method of the present invention, if a living being is detected in the vehicle interior and the vehicle subsequently receives a door lock signal, the computing unit executes emergency measures. These can include various actions, such as activating lighting devices, such as headlights, taillights, and turn signals, as well as emitting an acoustic warning signal, such as the vehicle's horn. This can alert passersby around the vehicle. Similarly, a warning message can be transmitted via a wireless communication connection, for example, to the vehicle owner's mobile terminal or to rescue services. Furthermore, the computing unit can control vehicle functions, such as opening power windows or activating the vehicle's air conditioning, for example, to prevent overheating in summer.

[0032] By performing emergency measures after the vehicle is locked, unnecessary initiation of corresponding emergency measures is prevented. If the vehicle doors are unlocked, this usually means that the driver of the vehicle is there or nearby. Therefore, there is basically no danger to children or animals inside the vehicle. In contrast, if the vehicle door When the vehicle is locked, there is no one to drive the vehicle, which increases the risk to children and / or animals left inside the vehicle.

[0033] According to the present invention, a vehicle is provided with at least two microphones and a computing unit, the microphones and the computing unit being configured to execute the aforementioned method. The vehicle may be any vehicle, such as a car, truck, transporter, or bus. The microphones are arranged at various positions in the vehicle interior, thereby enabling a particularly comprehensive and differentiated recording of various neural sound fields. The computing unit may be a central on-board computer, a control device of a vehicle subsystem, or a telematics unit. The various steps performed by the computing unit may also be distributed among several computing units physically separated from each other. One or more computing units may include computer-readable storage media, each of which may store a computer program product. When the computer program product is executed by a processor, the method steps described in connection with the method according to the present invention are realized in the vehicle.

[0034] Further advantageous embodiments of the method according to the invention for detecting living organisms in a vehicle interior as well as of the vehicle will become apparent from the exemplary embodiments which are described in detail below with reference to the drawings. [Brief explanation of the drawings]

[0035] [Figure 1] Schematic diagram of a residential neural sound field. [Figure 2] 1 is a schematic diagram of the steps of a method according to the present invention for identifying the presence of a living being in a vehicle cabin based on the generation of a neural acoustic field; DETAILED DESCRIPTION OF THE INVENTION

[0036] A neural acoustic field (NAF) is a representation of the acoustic properties of the measured environment. Andrew Luo et al. describe how to generate such a neural acoustic field in the paper Learning Neural Acoustic Fields, https: / / doi.org / 10.48550 / arXiv.2204.00628. Such a neural acoustic field can be used to estimate the structure, i.e., the geometry, of the examined space. This allows for the derivation of geometric information based purely on acoustic examination.

[0037] Fig. 1 shows such a nerve acoustic field NAF of a dwelling 6. The figure shows a sound source 7 placed anywhere within the surroundings of the object to be examined (here, the dwelling 6). An acoustic signal emitted from the sound source 7 is detected using an acoustic detection means, for example a microphone (stereo, if necessary). Sound waves 8 diffusing around the object to be measured are also shown. The solid lines represent the sound emitted from the sound source 7, and the dashed lines represent echoes reflected from the surroundings. The sound source 7 and the sound waves not shown in the figure acoustic The surroundings of the object to be inspected are mapped while varying the location of the detector. A spectrogram is generated by a short-time Fourier transform of the pulse response recorded by the acoustic detector. The spectrograms of a large number of individual measurements are then provided as inputs to a machine learning model, such as an artificial neural network. The artificial neural network then calculates the neural acoustic field (NAF), shown in Figure 1. Such a neural acoustic field (NAF) indicates the sound loudness or sound pressure level as a local distribution in the inspected surroundings, depending on the position of the used sound source 7. In Figure 1, the sound loudness is shown as a gray scale, with light gray corresponding to loud sounds and dark gray or black corresponding to soft sounds. Gaps 9 corresponding to walls or objects such as furniture are clearly visible.

[0038] This technique, described in the publication by Andrew Lou et al., can be applied to detect organisms in the vehicle interior. This is made possible by the method according to the present invention, the method steps of which are shown diagrammatically in Figure 2.

[0039] First, the vehicle's interior 2 needs to be suitably measured in a measurement campaign. For this purpose, different numbers of dummies with different specifications are placed in different seating positions in the vehicle interior 2, and different acoustic stimulus signals for measuring background noise are output into the vehicle interior 2. Next, a large number of neural acoustic fields NAFs are generated for different presence scenarios and collected in an NAF database 5. This NAF database 5 is then integrated into the vehicle's computing unit 4, which executes the method according to the invention.

[0040] When the vehicle is later used, the background noise at the time of closing the vehicle door is recorded by a plurality of microphones 3, here for example four microphones 3 distributed in the vehicle compartment 2, in order to identify the presence of a living being in the vehicle compartment 2. In the exemplary embodiment shown in Fig. 2, an adult who is located in the driver's seat of the vehicle and drives the vehicle, and a child who is located in the right rear seat, are present as living beings 1 in the vehicle compartment 2. The presence of the mentioned living beings 1 affects the diffusion of sound waves 8 in the vehicle compartment 2 when the vehicle door is closed. Therefore, the vehicle door is used as a sound source 7.

[0041] From the background noise thus recorded, the calculation unit 4 can derive an individual or averaged nerve acoustic field NAF for each microphone 3. This is shown in the top right of FIG. 2, which shows gaps 9 corresponding to the locations of structures in the vehicle cabin 2. Shown is a cross-section of the nerve acoustic field NAF at a specific geodetic height. This geodetic height corresponds in particular to the height at which each microphone 3 is mounted in the vehicle cabin 2. It is particularly advantageous if multiple microphones 3 are arranged at different geodetic heights in the vehicle cabin 2, since this allows cross-sections to be generated at various heights. In addition to the living organism 1, gaps 9 are also shown for the corresponding backrests of the vehicle seats.

[0042] This procedure, illustrated in FIG. 2, can be implemented explicitly or implicitly. In the explicit procedure, the neural acoustic fields NAFs derived from background noise are compared geometrically with the neural acoustic fields NAFs stored in the NAF database 5. That is, the correlation of overlapping lines is calculated, for example, mathematically. In contrast, in the implicit procedure, the respective neural acoustic fields NAFs are only checked for "similarity." For this purpose, artificial intelligence, in particular, can compare the respective data structures of the neural acoustic fields NAFs with each other. Such comparisons using corresponding machine learning models are performed as a black box.

[0043] To compare the background noise or the nerve sound fields NAF derived from the background noise with the NAF database 5, the calculation unit 4 checks which of the nerve sound fields NAF stored in the NAF database 5 has the highest match. For this purpose, the calculation unit 4 reads the corresponding NAF database 5. By way of example, different match probabilities are shown. At least the nerve sound field with the highest match is selected as the match in the result 10.

[0044] In response, the calculation unit 4 derives that the respective neural acoustic field NAF presence scenario that was present at the time of recording in the initial measurement campaign still applies to the vehicle. Thus, it is possible to calculate at least whether or not a living organism 1 is present in the vehicle compartment 2. The more clearly a particular neural acoustic field NAF in the NAF database 5 matches the neural acoustic field NAF derived for the background noise, the more information can be derived, in particular the number of living organisms 1, the type of living organism 1 (i.e., adult, child, animal, etc.), and in which seat the living organism 1 is located. [Prior art documents] [Patent documents]

[0045] [Patent Document 1] German Patent Application Publication No. 102019212412

Non-licensed literature

[0046]

Non-licensed literature 1

Claims

1. 1. A method for detecting a living being (1) in a vehicle cabin (2), wherein the vehicle cabin (2) is acoustically monitored by at least two microphones (3), and a computing unit (4) evaluates the signals generated by the microphones (3) using artificial intelligence, a) implementing a NAF database (5) in the computing unit (4), the NAF database (5) containing a number of NAFs of the vehicle cabin (2) for various presence scenarios, the NAF database (5) being generated by recording at least one NAF in each case during an initial measurement campaign in a representative vehicle cabin, varying at least the type of dummy, the number of dummy ... b) recording background noise in the vehicle compartment (2) with the microphone (3) at least when the vehicle door is closed; c) comparing the recorded background noise with the NAFs stored in the NAF database (5) by the calculation unit (4); d) calculating by said calculation unit (4) at least one nerve acoustic field (NAF) that has the highest match to said recorded background noise; e) determining the presence of a living being (1) in the vehicle cabin (2) by the computing unit (4) in response to each of the presence scenarios based on the neural acoustic fields (NAF) calculated in step d). A method characterized by:

2. The sound of a vehicle door slamming is used as the acoustic signal source for the initial measurement campaign to generate the NAF database (5), and in particular the nerve acoustic field (NAF) is also recorded with variations of the vehicle door slamming.

2. The method of claim 1.

3. The calculation unit (4) calculates which vehicle door is closed by taking into account a door closing sensor when the vehicle door is closed.

2. The method of claim 1.

4. Depending on the vehicle equipment, different initial measurement campaigns are performed to generate said NAF database (5).

2. The method of claim 1.

5. The computing unit (4) performs measurement campaigns to extend the NAF database (5) during the vehicle usage phase, and information describing the current presence scenario is actively queried by the user.

2. The method of claim 1.

6. Steps b) to e) are performed again after the vehicle door is closed, with the acoustic stimulus signal being output from at least one speaker disposed within the vehicle.

2. The method of claim 1.

7. The acoustic stimulus signal output from the at least one speaker includes only frequencies in the ultrasonic frequency band and / or the infrared frequency band.

7. The method according to claim 6.

8. The calculation unit (4) also uses artificial intelligence in step d) to identify the nerve acoustic field (NAF) that matches the background noise.

2. The method of claim 1.

9. When a living being (1) is detected in the vehicle compartment (2) and the vehicle subsequently receives a door lock signal, the computing unit (4) executes emergency measures.

2. The method of claim 1.

10. A vehicle comprising at least two microphones (3) and a calculation unit (4), The microphone (3) and the computing unit (4) are configured to perform the method according to any one of claims 1 to 9. A vehicle characterized by:

Images