DE102017213246A1 - Method, apparatus and computer program for generating auditory messages - Google Patents

Abstract

The invention describes a method, apparatus and a corresponding computer program for generating auditory reports of a computer carried by a vehicle (1), the auditory message comprising a voice message to at least one occupant (11, 12, 13) of the vehicle (1). The step of determining at least one content-related feature of a linguistic communication between the at least two occupants (11, 12, 13) of the vehicle (1) is carried out. Furthermore, the auditory message is adapted in dependence on the determined content characteristic of the linguistic communication between the occupants (11, 12, 13) of the vehicle (1).

Description

The invention relates to a method, a device and a corresponding computer program for generating auditory messages in the interior of a vehicle.

Modern vehicles have on-board computers and multimedia systems, via which the audio content of different audio sources in the interior of the vehicle can be played. In addition, the on-board computer or the multimedia system are able to output messages to the occupants of the vehicle by means of a voice output. Such messages, referred to in this specification as auditory reporting, may be e.g. Announcements of a navigation system, answers to voice commands or gestures, various hints (for example, to the environment, the weather, the condition of the vehicle, ...) and so on. There is also a temporary decrease in the volume of the multimedia contents being output, e.g. Music is known, while the on-board computer provides linguistic information, e.g. outputs a navigation information for the driver.

On the one hand there is a need to spend the largest possible amount of information in a vehicle to the occupants to meet the information needs. In addition, thereby the deflection of the occupants can be reduced by the observation or operation of displays of the vehicle and / or carried in the vehicle terminals.

On the other hand, there is typically a need or necessity for the occupants of the vehicle to communicate with each other. Such, in particular linguistic, communication can often have a very high relevance for users of the vehicle. The ability of the occupants to communicate with each other as undisturbed as possible may at times be much more important to the users of the vehicle than the technical parameters of the vehicle, such as e.g. engine performance, manufacturing quality, vehicle design or even the make or make of the vehicle. Such communication is nevertheless very often disturbed by a large number of voice messages issued in the vehicle. Therefore, there is a contradiction between the desire of the occupants, in particular the driver, to perceive the voice messages of the on-board computer and to communicate undisturbed the need of the same or other occupants while driving.

It is an object of the invention to provide a method, a device and a computer program which functionally improve the generation of auditory messages in a vehicle. In particular, a mutual compatibility of the linguistic communication between the inmates and auditory messages to be issued should be improved.

These objects are achieved by a method according to the features of claim 1, a computer program according to the features of claim 16 and a device according to the features of claim 17. Advantageous embodiments are specified in the dependent claims.

A method is proposed for generating auditory messages of a computer carried by a vehicle, wherein the auditory message comprises a voice message to at least one occupant of the vehicle.

Preferably, the auditory message comprises at least one linguistic message, which is generated by means of speech synthesis or from a database with prefabricated sounds, words, sentence sections or sentences. Auditory messages may, e.g. by means of an artificial intelligence, generated, preferably not prefabricated or predetermined, linguistic structures include. In a simplified case, the auditory message comprises information, e.g. a section of text that is audibly output in a preset language (computer or vehicle).

For example, e.g. the announcements of a direction indication of a navigation system, announcements of the on-board computer about the condition of the vehicle or current information about certain dangers are understood as auditory messages.

Auditory messages can be generated or output acoustically depending on a predetermined condition. For example, auditory messages, in particular linguistically, may include information to be output that depends on the driving context of the vehicle (eg traffic signs, traffic jams, other road users) and / or driver's condition (fatigue detection, distraction) and / or driver's actions (references to possible consequences of an action) are generated.

Furthermore, an automated reading-out of texts, for example incoming SMS, e-mails or entries (posts, tweaks, breaking news, etc.) in social networks and the like by means of the computer (on-board computer), can also be regarded as an auditory message.

An auditory message can also, at least predominantly, for a particular occupant, eg to the driver of the vehicle or to a specific one be determined further inmates (passenger). An auditory message may be a driver information item which relates, for example, to action recommendations, instructions, status messages in connection with at least partially automated driving and / or a maneuver that can be automated or performed by the driver, eg parking maneuvers of the vehicle. Auditory messages may be, for example, a human machine interface of a digital artificial intelligence based vehicle assistant and / or a "co-pilot". Such a "co-pilot" can also be designed to co-control the vehicle.

In addition, the term auditory messages also includes other, in particular interactively generated, voice messages that serve, for example, the entertainment of inmates. For example, an auditory message may contain a response, e.g. Information or replica to a question asked by an occupant, in particular the driver, or a gesture or facial expression made by the body language.

In particular, the auditory message relates to circumferences that are not limited to mere perception (listening, such as music) but to the interpretation of the corresponding message (understanding, distinguishing) by at least one occupant, e.g. the driver, are determined.

In the present description, a computer carried by the vehicle is understood to mean, in particular, an on-board computer of the vehicle. A computer carried along with the vehicle can also be an external unit of the vehicle, which is connected wirelessly or by wire to the vehicle, in particular to the on-board network of the vehicle. Such a unit is, for example, an in-vehicle user equipment, such as e.g. a portable navigation device, a mobile user device, e.g. a personal mobile radio terminal (e.g., a smartphone), data glasses, and the like. In particular, the output of the auditory message can take place by means of the multimedia system of the vehicle, in particular by the speakers installed in the vehicle. Preferably, the computer may also be part of the multimedia system of the vehicle, or part of a combined device.

In the method, at least one content-related feature of a linguistic communication between the occupants of the vehicle is determined. Linguistic communication is, for example, a conversation conducted in the interior of the vehicle between inmates. Such a conversation may be a monologue of an occupant or a dialogue between two or more occupants in the vehicle. Such linguistic communication between the occupants of the vehicle may be explicitly affected by further speech-related audio context in the vehicle, such as e.g. a music title, news program, a combined radio program or an audiobook. The linguistic communication also differs from a further, audible in the interior of the vehicle audible background as well as coming from outside the vehicle, but audible in the interior of the vehicle, acoustic content.

In the method, the auditory message is further adjusted as a function of the determined content characteristic of the linguistic communication between the occupants of the vehicle.

The at least one content-related feature of the linguistic communication between the occupants of the vehicle can be determined with one or more microphones present or carried in the interior of the vehicle. These may be built-in or additional microphones, e.g. act one or more belonging to a hands-free device of the vehicle or a mobile user device microphones. A content-related feature of the linguistic communication between at least two occupants can be detected, in particular recognized, in particular evaluated by means of signal processing, speech recognition, speech analysis and the like. The detection can take place with means of the on-board computer and / or a mobile user device carried along.

A linguistic communication can also be distinguished in the method from another conceivable audio context in the interior of the vehicle.

In other words, in the present method, a (at least simple) substantive evaluation or analysis of a linguistic communication taking place in the interior of the vehicle is carried out. The computer, which is carried by the vehicle, takes into account linguistic communication not directed or addressed to it, and then adjusts the auditory message which it is to generate. The auditory message to be adjusted is one which is preferably independent of the communication between the occupants.

The adjustment of the auditory message in the process can be based on an auditory message (eg, at least partially already generated as data) and / or on the (initial or repeated) generation of the auditory message (eg from an existing information for at least one occupant) consideration relate to the identified adjustment. In particular, the adaptation can be carried out such that in the method adaptation information, eg representing control signals, for controlling the computer in which the auditory message is generated, and / or for controlling the audio output device of the vehicle, eg the multimedia system, with which the auditory message is issued is determined. Such control signals may be determined within the vehicle or in another device (carried along with the vehicle or spaced apart, eg arranged in a cloud) and transmitted for carrying out the adaptation.

For example, adjusting the auditory message in the method may involve a content parameter (simplified: content) and / or a linguistic parameter of the auditory message. A content or linguistic parameter may e.g. represent the choice of words (e.g., a selection of two or more formulation options, voice, speech rate, sentence structure parameters, e.g., order of words or major and minor sentences of a linguistic message), and the like.

In the process, a content-related feature, in particular a textual content of the auditory message to be output, can be controlled depending on a content-related feature of the communication between at least two occupants, wherein the contents of the auditory message with the content of the communication between the inmates is not or at least not immediate or at least not causally, thematically or logically related.

In a further embodiment of the method, the, e.g. components of the communication message represented by the auditory message, which are known from communications science, vary, in particular in a targeted manner. For example, the components of "appeal", "relationship", "self-disclosure", "content" (pure information content) known from communications science can be adapted to a message of the auditory message to be issued, in particular to be generated. For example, the proportions between these components can be adjusted.

A parameter of the auditory message to be adjusted may represent a parameter of the communication behavior. Such a parameter may determine a communication behavior of the computer to at least one occupant of the vehicle. Such a parameter may e.g. when generating the auditory messages for communication with at least one occupant corresponding to the applied communication model, in particular with regard to the interpretation of the voice messages of the occupant and / or with regard to generating a response. For example, in the method, a parameter that is a communication behavior based on or in accordance with a model based on a transaction analysis or an SR model (= stimulus-response model) and / or a predetermined communication target can be changed, in particular adapted.

For example, depending on at least one parameter of the communication in the interior of the vehicle, the communication behavior of the on-board computer can be adjusted approximately according to the logic or contexts known from a transaction analysis or an SR model.

Alternatively or additionally, adjusting the auditory message may include adjusting, e.g. Controlling or regulating a parameter of the audio signal generated when generating or outputting the auditory message, in particular as regards acoustic parameters.

The adaptation of the at least one auditory message may be performed in the method as an optimization according to one or more predetermined criteria, e.g. the user settings.

The method may be applied, in particular simultaneously, to multiple linguistic communications in the vehicle (e.g., between any two or more occupants). For example, it can be applied to different concurrent communications. As different communications, for example, communications may be viewed that are conducted between different people and / or directed to different people. Although communication is referred to in the context of this description, it may mean one, in particular one, of a plurality of or possible communications in the vehicle, e.g. a communication that concerns the driver.

According to an expedient embodiment, the at least one content-related feature of the linguistic communication is determined by means of a linguistic analysis, wherein in the linguistic analysis one or more of the following features of the linguistic communication are processed. In this case, a semantics of the linguistic communication and / or a topic or a topic feature of the linguistic communication and / or an information as to whether or not the linguistic communication comprises a content-contained message can be processed. One content Completed message in the sense of the present description may be a sentence or phrase.

The topic features can be determined in the method by means of a stored association rule, association logic, a database representing different strong (association) associations between the topics, words, typical sentence structures, etc.

If, for example, it is recognized, in particular predicated, that or when an occupant ("speaker") has spoken or is about to finish his sentence and / or a phrase and / or a message, then the phrase that points to the end of the phrase or phrase can be spoken the message adapted to the message generated or issued.

Preferably, it can be determined from the content-related feature at which time a linguistic message of the speaking occupant of the vehicle will be completed. Thereupon, for example, the beginning of at least part of the auditory report can be planned. In the case of incomplete sentences, the computer waits a little longer to continue playing the adjusted or unmatched auditory message, assuming that more speech content will follow from the talking occupant.

In the method, a comparison can preferably be made between a physical, in particular linguistic, analysis applied to the linguistic communication of the at least two inmates and an evaluation of one or more corporeal impulses (occurring in the time interval of approximately 1, 3, 5, 10 seconds) Mimic expressions, for example, the facial muscles or lip movements of the occupant or those who are detected with an indoor camera, carried out and taken into account.

A further expedient embodiment provides that the content of the auditory message is adapted as a function of the determined content characteristic of the linguistic communication between at least two occupants of the vehicle. In this example, adaptation of one or more information represented by the auditory message, e.g. the acoustically output text.

In the method, an adaptation of the auditory message depending on the subject of the speech communication between the occupants, e.g. can be determined according to predetermined criteria or logic. For example, a different adjustment of the auditory message may be provided, depending on whether or not the occupants are talking about a topic related to the vehicle, the driving context, the route, etc.

According to this refinement, depending on the determined content-related feature of the linguistic communication between the occupants of the vehicle, a particularly forward-looking, content-related adaptation of the auditory message is carried out. For example, depending on the determined, in particular predicated, content characteristics of the linguistic communication between the occupants of the vehicle, a reduction of the contents of the auditory message and / or addition of further contents for auditory reporting and / or a more detailed description of the contents of the auditory message can take place. The computer thus controls the contents of its generated auditory message depending on a content-related feature, but the contents of the auditory message with the communication of the inmates are not directly related, especially not thematically, logically or causally.

In the method, an auditory message, which also includes a part of the information not represented by speech, can be adapted. For example, such a sound accompanying the at least one auditory message can be adapted.

According to a further expedient embodiment, a time for the output of at least part of the auditory message is determined as a function of the determined content-related feature. The content characteristic can be predicated. Depending on the content-related feature, it can be determined, for example, that the time for the output takes place only after the determination of an end of the linguistic communication. Alternatively, it may be determined that the at least part of the auditory message occurs in a determined pause of the linguistic communication.

A further expedient embodiment provides that the auditory message is paused and / or divided into at least two parts as a function of the determined content-related feature, the output of which is effected, for example, with a (temporal) interruption and / or whose sequence is changed or changeable. In this embodiment variant too, the content-related feature of the linguistic communication can comprise a linguistic parameter. Depending on this, it may be determined to interrupt the auditory message and the rest of the auditory message at a later time, e.g. after the completion of the linguistic communication.

A further expedient embodiment provides that the auditory message depends on the identified content-related feature stylistically, with regard to their voice and / or their pitch and / or undertone.

The pitch of the auditory message and / or the content of the auditory message may be adjusted such that e.g. a more or less polite, preferably stylistically fitting insertion into the linguistic communication of the occupants of the vehicle takes place. Depending on the content-related parameter of the linguistic communication, the auditory message can be varied in word choice, tone, voice or linguistic parameter. These may be adapted or adapted according to a communication model, e.g. a model based on a transaction analysis or a so-called SR (Stimulus Response) model.

In a simplified case, the computer may adapt auditory messages generated based on the linguistic communication content characteristic determined. The computer can then be e.g. output the auditory message in a "quiet" tone, or output it in a "clear" tone, or vary a portion of empathetic expressions, or choose a communication style according to a transaction model. The above examples can also be combined in any way.

Alternatively or additionally, the auditory message regarding their voice and / or their pitch and / or an undertone (also) depending on the voice and / or pitch and / or undertone of the linguistic communication between at least two inmates can be adjusted. In this case, the method may include an evaluation of the voice and / or the pitch and / or undertone of the linguistic communication between at least two occupants.

The adaptation of the style of the auditory message to be output, in particular with regard to their voice and / or their pitch and / or undertone, may also relate to contents (content characteristics of the messages in question) that do not, in particular not directly, with the contents of the communication between the inmates , in particular not causal, logical, thematically related.

A further expedient embodiment provides that recognition of one or more keywords in the communication of at least one occupant and / or recognition of the gesture of at least one occupant is performed and then the auditory message regarding a content feature and / or an acoustic parameter (with the this is generated or output) is adjusted. For example, as an indication for the adjustment of the auditory message, a mention of the target person by name and / or predefined keywords such as "to all", "all listen to", "children", "behind her" can be used. In this context, it is useful if keywords such as those mentioned above are defined in advance and learned. Likewise, it is convenient to define in advance and to learn in advance given gestures of the occupants that characterize a direction of the communication.

In addition, recognition of predefined, in particular learned, key words and / or gestures of the occupants which characterize a direction of the communication can take place, and subsequently an adaptation of the auditory message, preferably of at least one surround parameter, depending on the recognized keywords and / or gestures.

A further expedient embodiment provides that by means of a linguistic evaluation of the linguistic communication, a prediction of a temporal parameter, in particular a chronological progression, of the linguistic communication is carried out and the temporal parameter is processed during the generation and / or the output of the auditory message. The consideration of the temporal parameter can take place with regard to an initial and / or an end and / or the chronological progression of the amplitudes and / or frequency parameters. The linguistic analysis can be done in terms of sentences or phrases. When generating the auditory report, in particular with regard to its content, its division or the time of issue, the linguistic analysis can be taken into account. Expediently, a forward-looking adaptation of the content to be output as an auditory message is carried out. Alternatively or additionally, the prediction of the time course of the voice of the speaking occupant can also take place by means of the analysis of the bodily movements, his facial expressions, a lip movement pattern and / or the body language.

Particularly preferably, at least two temporal parameters are determined, in particular predicated, characterizing a time interval in which the auditory message to be output can be output in the interior of the vehicle, in particular with no, no significant or at least reduced temporal overlap. It is particularly preferable then to perform a predictive adaptation of the audio signal to be output, e.g. in the form of scheduling the output of one or more portions of the audio signal.

Depending on one or more temporal parameters, the starting of the output of a (specific) section of the audio signal, and / or the end of the output of a ( determined portion of the auditory message, and / or the time course of the amplitudes and / or frequency parameters of the auditory message, adjusted. Such an adaptation can be carried out cyclically, continuously in the method.

A further expedient embodiment provides that the type and / or content of another audible audio context which can be heard in the interior of the vehicle, deviating in particular from linguistically represented information, is determined and processed as a function of an analysis result for adapting the auditory message to be output.

This embodiment provides that an adaptation of the auditory message depends on the type of one or more further audible audio contexts audible in the vehicle. The type of audio context can be divided into the following or similar types: music titles; Language program; Telecommunication content (telephone call, read SMS, messenger, etc.). The adaptation of the auditory message can be done depending on the content of this further audio context in the vehicle or the vehicle.

The content of the, in particular in the interior audible, in particular currently existing or output, audio context can in the process, e.g. acoustically, as an audio signal to be output or data to be output, are recorded and subdivided according to the following or comparable classes: start / end of a sentence and / or phrase of a multimedia program comprising multimedia (radio program); News, breaking news, entertainment; Voice messaging, telecommunications content (e.g., a hands-free call, with or without hands-free from a passenger, a telephone or video conference, and the like).

For example, depending on determining a high priority of a current or predicted further audio context, e.g. as "important" and / or "urgent", lead to a much greater adaptation of the auditory message than in an audio context with a medium or lower priority. For example, upon detecting a temporal overlap between an audio context of a particular type and / or content and an auditory message, the auditory message may be output at an earlier or later time and / or in a shortened form and / or with a changed acoustic parameter.

The content feature of the linguistic communication may be the importance and / or urgency of the communication content and / or a particular word choice and / or a particular gesture accompanying or preceded by the auditory message. Accordingly, according to this embodiment, an audio signal to be outputted from a song or an entertainment program, especially with regard to its acoustic effect to be generated in the vehicle, can be adjusted if it is also determined that one or more occupants, e.g. to communicate an important and / or urgent message on the basis of facial expressions, volume, wording and the like. Depending on the recognized type and / or the content of further, in the interior of the vehicle audible audio context, a linguistic communication of an occupant of the vehicle of these other, in particular a language or text communication comprising content can be distinguished.

In addition, a parameter may be further determined and processed in the interior of the vehicle audible audio context. The adjustment of the parameter of the audio signal may be made depending on the kind of the audio context. It can be discriminated between an audio context originating in the interior of the vehicle and an audible signal coming from and entering from outside the vehicle interior.

It may also be between an audio context (e.g., a radio broadcast, a talk show, etc.) output from a vehicle-mounted playback device (e.g., a user device with a speaker) and another audio context, e.g. from a background noise, can be distinguished. Furthermore, a telephone call conducted by at least one occupant communication in the process can also be considered as a type of linguistic communication. This may lead to a similar or different adaptation of the auditory report to be issued.

Another expedient embodiment provides that, depending on an evaluation of at least one physical attitude and / or a gesture of the at least one occupant of the vehicle is detected that a linguistic message explicitly to the entrained in the vehicle computer or explicitly not to the computer of the vehicle and / or is directed to a specific occupant of the vehicle.

This embodiment allows the recognition that the speech communication is explicitly not addressed to the on-board computer, e.g. when a speaker says one sentence and turns physically to another occupant of the vehicle. Thereafter, the linguistic message output by the speaker is not or at least not analyzed to generate a response or command execution.

Furthermore, in the method, a determination and consideration of a (scientific, mathematical) relationship between a parameter of one or more physical impulses of at least one occupant, which are dependent on certain patterns of gestures and / or facial expressions, and a concomitant or subsequent content-related feature of the linguistic communication , in particular the speaking of the occupant or a reaction of the occupant to the speech of another occupant or to a specific auditory message.

It is also possible to avoid unwanted answers or unwanted (textual) analysis of the conversation. The evaluation is then rather purposive, in particular only to adapt the parameter of the auditory messages generated by the computer. For example, depending on a direction parameter of the voice communication between at least two occupants of the vehicle, the sound source corresponding to the generated voice of the on-board computer may be virtually moved, removed, brought closer or moved. For this purpose, the physical posture and / or a gesture of at least one occupant of the vehicle can be detected with a camera. A donation of the head of an occupant or a body rotation and / or eye contact to another occupant can be recognized thereby.

A further expedient embodiment provides that a probability measure for a recognized linguistic message of an occupant is determined, wherein when the probability measure exceeds a threshold that the message is not directed to the computer, in particular with the expectation of a response, the auditory Message is adjusted.

Preferably, in the method, a probability measure can be determined and taken into account that at least part of the auditory communication, in particular a specific (audible in the interior of the vehicle) linguistic message is not directed to the on-board computer and / or directed to another listener and / or addressed to a listener connected by mobile. Thereupon, the adaptation of the at least one parameter of the audio content can take place taking into account one or more determined probability measures.

Inmissions of passengers of the vehicle are analyzed to see whether and, if so, with what probability they are directed to the computer (on-board computer) or not. If the determination of a determined probability measure does not exceed a predetermined threshold value, then the computer should not analyze the linguistic message, not answer it or not (perceptible for the occupant), in particular not process in the sense of a command interpretation or execute a voice command. This can be achieved, for example, taking into account a content-related, in particular linguistic, speech analysis.

A further expedient refinement provides that at least one feature of the linguistic communication between the at least two occupants of the vehicle is determined as a function of the recognized body language and / or a voice position, in particular a change in voice, of at least one of the occupants and / or to adapt the generation and / or or the output of the auditory message is processed.

A further expedient embodiment provides that an order of at least two auditory messages and / or individual contents of a respective auditory message changes depending on a current communication or a predicted communication project and / or a vote of at least one occupant and / or occupant at a particular seat becomes. For example, the adjustment of the auditory messages to be output can take place in such a way that the data sections which belong to different auditory messages and / or individual parts of a (split) auditory communication are output in a changed order.

Expediently, an order is selected in which the content to be output in the form of the auditory message must be broken down into as few parts as possible and / or a length of the pauses between the linguistic communication of the occupants in the vehicle and at least part of the auditory message is maintained.

A further expedient embodiment provides that a content-related feature of the linguistic communication between at least two occupants of the vehicle is determined and / or the type and / or priority of one or more auditory messages to be output is determined. Thereupon, an order of the at least two auditory messages and / or individual contents of the auditory message, for example the two parts of a divided auditory message, is determined, depending on the determined content characteristic of the linguistic communication and / or the determined type and / or the priority of the at least two auditory reports to be issued.

In this case, a targeted analysis and prioritization of multiple auditory messages to be audibly issued or multiple portions of an auditory message may be performed. Thereupon two can or several auditory messages or parts of a split auditory report are placed on a timeline as part of a schedule.

Preferably, in the method, an analysis of the type and / or priority of at least two auditory reports to be output or parts of an auditory report can be carried out. Thereupon, the output of at least two auditory messages or sections of auditory reporting may be scheduled, in particular depending on the differences between their types and / or priorities, particularly particular time intervals and / or one or more expected criteria or events (eg, achievement a certain road section, for example, before an intersection, exit, etc.) are assigned.

Alternatively or additionally, in the method, a comparison of the parameters of the audio context, representing a type and / or a predetermined or determined from the type priority measure, executed, and the parameter of an auditory message to be issued, a kind and / or a predetermined or one of the kind determined priority measure represents the auditory message executed. Such an adjustment can be carried out, for example, according to a prescribed rule, logic, assignment table, etc. Depending on the adjustment of the parameters, the acoustic message to be output can be adapted.

At least one optimization criterion for adapting at least one auditory message preferably represents a sufficient audibility and / or perceptibility and / or (acoustic) intelligibility of at least one (specific) auditory communication between the occupants and / or the at least one auditory message. Such perceptibility or audibility or intelligibility can be determined, for example, using a psychoacoustic model of a human acoustic perception and taken into account in the method.

At least one further optimization criterion for adapting at least one auditory message preferably represents a minimization of a temporal overlap between the speech of at least one occupant and the output of the auditory message.

Another optimization criterion may be at least one vehicle setting or user setting, e.g. can be assigned to a seat or a user profile of an occupant. This can represent, for example, the extent to which auditory communication, in particular from a certain direction and / or to what extent one, in particular a specific, auditory message should be audible. Certain perception characteristics of at least one particular person, e.g. Parameter representing an audiogram of a particular occupant, be taken into account.

Consequently, the method may be carried out such that both at least one auditory communication and at least one auditory message, in particular from a particular occupant, is sufficiently audible and / or to a desired extent.

A further expedient embodiment provides that it is determined that a part of the auditory message is not sufficiently perceptible with a probability exceeding a predetermined threshold, and the part of the auditory message is repeated. In this embodiment, at least a portion of the auditory message may be repeated such that the repeated portion of the auditory message is a valid, i. sufficiently complete and comprehensible to the inmates.

Alternatively or additionally, an adaptation of the audio signal corresponding to the auditory message can be carried out such that the signal energy of at least a part of the auditory message is distributed to one or more frequency ranges which have a lesser, in particular the smallest possible overlap or correlation with the frequency ranges of the human voice. in particular to the voice of a particular occupant (the "speaker"). In this case, the perceptibility and / or comprehensibility of the linguistic communication between the occupants of the vehicle and / or the auditory message can be increased or a mutual interference, in particular a masking (known, for example, from psychoacoustics) can be reduced.

In another example of the method, the computer, in particular an application corresponding to a "vehicle assistant" or a "co-pilot", may be configured to behave as a "cooperative" party in the interior of the vehicle, which is the communication between the occupants in the vehicle not adversely affected. In doing so, the computer may be provided with its own predetermined voice priority, which perceives the communication of the occupants in the vehicle and chooses an appropriate time in the communication to issue the auditory message. If it can be assumed that the auditory message has not been sufficiently heard, the computer can repeat it without any action on the part of the occupants being necessary.

In applying the method to at least two concurrent communications, the resulting adaptation of the content and / or audio signal parameter of an auditory message to be output may be made taking into account the at least two communications in the interior of the vehicle. In this case, the at least two different communications in the method can be taken into account, for example, with different predetermined criteria and / or with different influence on the at least one auditory message.

Preferably, in the method, a control of a temporal parameter of at least a part of the auditory message, a corresponding acoustic parameter, in particular a surround sound parameter and the content (content parameter) of the auditory message is dependent on one or more content characteristics of one or more linguistic communications in the interior of a vehicle between the at least two occupants of the vehicle.

In the method, at least two content-related features of the linguistic communication can be determined and taken into account, which include a, e.g. time-ordered sequence of several corresponding parameter values, e.g. with regard to the near past and / or about the current time and / or a prediction for the near future. For example, a parameter value in terms of content may also represent a measure of the change of a corresponding value over time.

The content characteristic does not have to correspond to the content of the communication. This may include only one or more parameter values representing one or more features of the communication (without reflecting the communication itself). In a particular example of the method, a linguistic communication feature may include, for example, an objective and / or subjective truth content of a corresponding portion of the linguistic communication (e.g., estimated in the method), e.g. a statement, represent.

At least one content-related feature of the linguistic communication can be used in the method on the basis of sensor data, a history, e.g. in connection with one or more communications in the vehicle interior or communications involving certain persons and / or considering a probability calculation. For example, at least one content-related feature of a current or forthcoming linguistic communication can be predicted by means of a comparison or a clearing of a multiplicity of content-related features and / or further parameters from the history. Such alignment can be performed in the method e.g. be executed within a backend.

In the method, it is preferable to adapt, in particular continuously or cyclically, to adapt at least one auditory message, e.g. a parameter of the auditory message. The adaptation can take place before and / or during the output of at least one section of the auditory message, depending on at least one of the past or currently determined content-related parameters and / or at least one predefined content-related parameter of the communication. The "adaptation" in the method means, in particular, a suitable modification, which serves, for example, to achieve one or more predefined goals, one or more predefined logical or mathematical relationships, in particular according to one or more optimization criteria.

Optionally, the computer may also be configured to capture, interprete a spoken language into a voice input, command execution or generation of a response. This results in synergistic advantages, because the (oriented to the speaking with the driver) features of such a computer for executing the method, in particular for the analysis of not directed to the on-board computer linguistic communication, can be shared or enabled.

In the context of this description, the vehicle is preferably a motor vehicle, in particular a passenger car (= passenger-power car). This is advantageous because the interior of such a vehicle forms a comparatively tight, acoustically largely isolated space in which multiple auditory communications and auditory messages are to be generated and heard within a confined space. In principle, however, the method can also be applied to other types of vehicles, e.g. a bus, train, watercraft, aircraft, if necessary, be applied in a mutatis mutandis adapted form. Thus, the visual perception of the driver and the need to accommodate the view of the surroundings of the vehicle on a display instrument (usually often), can be spared.

Furthermore, an application of the invention to an automated moving vehicle is particularly advantageous because the driver of such a vehicle due to the relief of driving tasks close the eyes for a long time or avert from driving. The auditory communication in general and in particular by the driver with another occupant and / or with the Computer becomes correspondingly more significant. The invention makes it much more comfortable to design and expand if necessary. Thus, for example, important information can also be output as generated language.

The invention further provides a computer program product that can be loaded directly into the internal memory of a digital computer and includes software code portions that perform the steps of the method set forth in this specification when the product is run on a computer. The computer program product can be realized on a data carrier, such as a DVD, a CD-ROM, a USB stick and the like. Likewise, the computer program product may be implemented as a program to be loaded that may be loaded over a network or wirelessly. The computer program product can also be designed as an update of a previous computer program, which includes, for example as part of a functional extension, for example in the context of a so-called "remote software update", the corresponding program code or software code sections relating to the execution of the method.
The invention further proposes an apparatus for generating auditory messages of a computer carried by a vehicle, the auditory message comprising a voice message to occupants of the vehicle. The device comprises a first means for determining at least one content characteristic of a linguistic communication between the occupants of the vehicle and a second means for adjusting the auditory message in dependence on the determined content characteristic of the linguistic communication between the occupants of the vehicle.

Preferably, the device may be implemented together with or based on a device for detecting, interpreting a spoken language for voice input, executing commands, or generating a response. In this case, the means (designed for speaking to the driver) of the device can also be set up for executing the method, in particular for analyzing the linguistic communication not directed to the on-board computer and / or adaptation of the auditory messages. Furthermore, the device may have further, e.g. include means carried or installed in the vehicle.

The device according to the invention has the same advantages as described above in connection with the method according to the invention.

In an expedient refinement, the device according to the invention has further means for carrying out the method according to the above-described and following description.

• 1 eine schematische Darstellung eines Fahrzeugs mit einer Vorrichtung zum adaptiven Erzeugen auditiver Meldungen.

The invention will be explained in more detail below with reference to embodiments in the drawing, without limiting the generality. Show it:

• 1 a schematic representation of a vehicle with an apparatus for adaptively generating auditory messages.

1 shows a schematic representation of a plan view of a vehicle 1 , In the vehicle 1 it is a motor vehicle, in particular a passenger car, a van or a minibus. The vehicle 1 can be powered by an internal combustion engine and / or an electric motor. The vehicle 1 can also be designed to be moved purely electrically. Optionally, the vehicle 1 have semi-autonomous or fully autonomous driving functions. Basically, the procedure can also apply to other types of vehicles 1 (eg bus, train, watercraft, aircraft) are applied in a suitably adapted form.

In the in 1 shown representation corresponds to the direction of travel of the vehicle 1 the arrow direction of the coordinate system marked "front". The vehicle 1 thus moves in the drawing in leaf level from bottom to top.

The vehicle 1 has only four seats for example. A left front seat is with 2L, a right front seat with 2R , a left, rear seat with 3L and a right, rear seat with 3R characterized. By way of example only sitting on the left front seat 2L , in the right front seat 2R and on the right, rear seat 3R one inmate each 11 . 12 . 13 , For example only, the left, rear seat 3L not used. The inmate 11 represents the driver of the vehicle 1 Only for illustration purposes is therefore in the direction of travel in front of the occupant 11 a steering wheel 14 or another control arranged to influence the lateral guidance of the vehicle.

The schematically illustrated vehicle 1 includes a multimedia system 4 , The output of auditory messages is the multimedia system 4 with a number of speakers 8th - 1 , ..., 8th - 7 connected. The speaker 8th - 1 , ..., 8th - 7 are distributed in the interior of the vehicle, with only the example of the speakers 8th - 1 centrally located in the area of the dashboard in front of the occupants. The speaker 8th - 2 . 8th - 4 . 8th - 6 are in the left area of the body of the vehicle and the speakers 8th - 4 . 8th - 5 . 8th - 7 in the right area of the body of the vehicle 1 arranged. The multimedia system 4 Optionally has a not shown Control unit, which is capable of a surround sound in the interior of the vehicle 1 to create. This means over the speakers 8th - 1 , ..., 8th - 7 output auditory information can be controlled by targeted control of Raumklangparametern, frequency parameters, amplitude curve and / or phase parameters so that the volume and / or sound properties can be formed differently at different seats of the vehicle.

As an alternative or in addition to the illustrated multimedia system, at least part of the method, in particular the output of the auditory message, can be carried out with a device carried in the vehicle, which is coupled to the vehicle at least for energy supply and / or data exchange for executing the method.

The vehicle 1 also has a computer on an on-board computer 5 which represents a computing unit. With the on-board computer 5 are a camera 6 for detecting the interior area of the vehicle and a microphone 7 for the reception of, by the inmates 11 . 12 . 13 arranged messages. As a computer could also be carried in the vehicle and not shown in the figure, in particular mobile terminal serve, such as a portable navigation system or a mobile terminal. These can use a wireless (eg Bluetooth) or a contact connection bidirectional data with the car, eg the multimedia system 4 , change. A variety of different auditory messages (information about listening and understanding) can be issued depending on predefined criteria or generated freely.

With the reference number 9 In general, a sensor is characterized. The sensors 9 For example, sensors for detecting a seat occupancy and / or sensors for detecting an applied seat belt and / or door sensors, which detect the opening and closing of a door may comprise. With the help of sensors 9 It is possible to change the occupancy of each seat 2L . 2R . 3L . 3R computer-aided to determine. In addition, to detect whether a seat is occupied, the camera 6 captured image of the interior to be evaluated. If, for example, it is determined that nobody is sitting on a certain seat of the vehicle, then it can be ruled out that a linguistic communication from the corresponding spatial direction is to be expected. In contrast, a better adaptation of the parameters of the audio content for at least one other spatial direction can take place.

The described device is used to output auditory messages, as well as other audio content other than the auditory messages, in such a way as to provide linguistic communication to the occupants 11 . 12 . 13 does not bother.

To do so, it is determined by means of the sensors and / or the computer (i.e., the on-board computer or the vehicle-mounted but vehicle-external terminal) when a suitable time for issuing the auditory message is. The device is also used to output the auditory messages such that they are intelligently controlled depending on the linguistic communication and / or content of the linguistic communication in the vehicle. For this purpose, the auditory message can be divided into several sections and output at different times, which take into account the flow of speech of the occupants in their linguistic communication. It can be a temporal shift, preference or a distribution of auditory notifications. It can also be an optimized placement of the disassembled parts of the auditory message regarding the order of their issue in the flow of communication within the vehicle.

The timing of the output of the auditory message is determined depending on a predicted acoustic content of the voice communication. In this case, a distinction is made between an interior generated and an exterior audible but audible interior acoustic content. In addition, a distinction between an acoustic content of the speech communication and an on-board or connected to the vehicle or carried in the vehicle audio output system can be done.

Particularly preferred are one or more temporal intervals, e.g. represented by a time and / or duration, determined which are suitable for outputting at least one (specific) auditory report or section of an auditory report and / or one or more sections of an auditory report having (respectively) a particular time duration; and / or are unsuitable.

The determination of the corresponding temporal parameter for the output of one or more auditory reports or the sections of the auditory messages may be suitable and / or unsuitable in the method from the determined time intervals and the information as to whether these more or less, in particular in a quantitative measure Output of the auditory messages are to be determined. In a simplified case, the appropriate time points or intervals can be determined from the temporal parameters (in an analogous manner) which indicate time points or time intervals which are not suitable for outputting the auditory messages.

In addition, a distinction between a type of audible in the interior audible content (eg music, audio program, voice communication) is possible. Different occupants or occupants in different seats may each have a different influence on the determination of the time for the output of the auditory message and / or on the substantive influence on the auditory message assigned. For example, when issuing the auditory alert, the computer may consider, for example, the driver's communication with another occupant regarding the time the auditory message was issued, while one between the occupant 12 in the passenger seat and the occupant 13 is considered to be irrelevant in the back seat on the right. This may be the case, for example, in the issuance of navigation announcements, which are usually only to the driver (ie occupant 11 ) of the vehicle 1 applies.

Optionally, in the method, a recognition of occupied seats in the vehicle can take place and a number of occupants present in the vehicle and a distribution of certain occupants via seats can be determined and taken into account. In this case, at least one potential listener of the speech communication and / or the auditory message in the interior of the vehicle can be determined. Furthermore, the adaptation of the auditory message can be carried out taking into account the at least one potential listener, in particular a sitting position of the potential listener. If, on the other hand, it is determined that there is nobody in a certain part of the vehicle interior, then an adaptation or optimization for the relevant part of the interior can be dispensed with and / or an adaptation of the adaptation optimized for at least one further part of the interior (eg seats) at least one auditory message can be executed.

Optionally, furthermore, an adaptation of the at least one parameter of the auditory message, an assignment of certain contents to be output and / or signal components of the at least one auditory message to at least one loudspeaker and / or by adjusting the balance between at least two, preferably between at least four, in the vehicle interior arranged loudspeakers take place.

Optionally, in the method, for example by means of a corresponding determination or modification of the one or more surround sound parameters, one or more (also virtual, audible at least by human perception in the interior of the vehicle) sound sources can be generated.

For example, a sound source which corresponds to the output of an auditory message, in particular with the one or more auditory messages, may each have different directions, distances, positions, acoustic parameters which can be controlled or adapted in particular in the method.

Generating the sound sources to one or more auditory messages may each refer to a particular interior region of the vehicle from which that sound source or sources are made perceptible (audible).

Preferably, in the method, a surround sound parameter representing a direction, distance, position, acoustic parameters to at least one sound source (perceptible from a specific seating position and / or head orientation of the occupant) can be determined and a corresponding sound source generated.

For example, auditory perceptible sound sources are generated for at least two different spatial areas of the interior of the vehicle or for two specific occupants or occupants at certain seats, based on an auditory message to be output, depending on said auditory communication parameter. In this case, these sound sources or the auditory messages corresponding to the sound sources may have different perceptible parameters, each suitably adapted, objectively measurable or subjectively perceived by the actually present or assumed occupants.

Optionally, depending on the recognition of the gender corresponding to the at least one voice with which said communication takes place in the interior of the vehicle, the gender of the voice in which a voice comprising an auditory message is generated can be determined. This can then be determined according to predetermined criteria, e.g. User settings or logic, each corresponding to the same or a different gender.

For example, if it is detected that there is an indoor communication with one or more male voices, then a voice message of the computer is output in a female voice and / or if it is detected that communication in the living space occurs with one or more female voices, then becomes a to be transmitted with the audio signal Voice message of the computer output in a male voice.

Optionally, depending on a content-related feature of the linguistic communication, the priority of the respective communication spoken by an occupant can be automatically estimated or taken into account. Thus, a traffic-relevant linguistic communication between the occupants can be prioritized, so that e.g. stopped an auditory message from the computer and at a later time, e.g. after completion of the linguistic communication between the inmates, is issued.

Preferably, a parameter of the linguistic communication between the occupants, e.g. An indicator of how long the communication will take, predicted and taken into account when adjusting the auditory message.

Depending on the nature of a determined or predicted acoustic content of the linguistic communication, a scheduled or currently issued auditory message may be paused and played back after the end of the linguistic communication (or sentence within the linguistic communication). For this purpose, the auditory message is divided into at least two parts.

The auditory message can be paused depending on the determined or predicted feature of the linguistic communication, in particular depending on the type of acoustic content. Alternatively or additionally, the content of the auditory message may be divided into at least two parts. Such a procedure may be initiated depending on the predicated auditory message, eg, before or during its issue. For example, the auditory message, such as an announcement of the onboard computer 5 paused or anticipatory, ie before its release, divided into two or more parts.

Will at auditory announcement issue by means of sensors 9 determined that at least one, in particular relevant, part of the auditory message was not sufficiently perceived by the occupants with a given probability, at least one relevant part (the part and / or a part related to the part) of the auditory message is repeated , The repetition preferably takes place in such a way that the repeated part of the auditory message results in a valid, ie sufficiently complete, sentence.

This may be a repetition of a portion of the auditory message, which is or has been significantly influenced by the determined auditory (linguistic) communication in the interior. In this case, e.g. an auditory message already started which has been substantially drowned out during its output by the auditory communication (strong and / or during a relatively long time interval) will be issued repeatedly.

As a repetition is also to be understood that a new auditory message or a portion of the auditory message is generated again, whereby at least one, not sufficiently transmitted content is thus compensated, supplemented or replaced.

In the method, linguistic communication (speech) can also be (explicitly) distinguished from another conceivable audio context in the interior of the vehicle, in particular also from a multimedia program comprising a language. In this case, a separation of a, for example, acoustically detected, signal in a, a linguistic communication comprehensive, part and / or a further linguistic content comprehensive and / or no language corresponding part done.

Linguistic communication in the interior of the vehicle may be affected by other audio contexts, e.g. of the multimedia system, preferably a comparison being carried out between an audio signal output from the multimedia system and a signal recorded with at least one microphone in the interior of the vehicle. In this case, for example, the linguistic contents (voice of a newscaster) output from the multimedia system can be distinguished from the voice of at least one occupant. The comparison can be determined by analysis of similarity, correlation, etc.

Furthermore, the further contents to be output from the multimedia system can be predicted and taken into account when adapting the at least one auditory message. For example, the content of a stream of stream or content read from a volume may be analyzed before it is rendered. Typically, such a data stream (read from the Internet, cloud or CD) is present at least a few seconds prior to its playback in a corresponding device (computer, memory or memory area). In the invention, it can also be analyzed (at least roughly, for example according to the amplitude curve).

In this case, the auditory messages can also be adapted taking into account a predicted audio context for the near future. In other words, the method can be used to look into the near future, in order then to determine the auditory messages with regard to their content parameters, parameters of the corresponding acoustic signal, with regard to the time of the output or the Order of at least two acoustic messages or parts of a, for example, split, acoustic message adapt.

In particular, at least the auditory message is adjusted, in particular with regard to its content feature, depending on (in the last few seconds) predicted parameters and at least one further adaptation, in particular with regard to an acoustic parameter, almost in real time. It is also possible to carry out a cyclical, ongoing adaptation of the auditory message, in particular almost in real time.

Nevertheless, a feature of content may also be generated during the generation of the auditory message to be output, e.g. deviating from a previously planned content-related characteristic, (ongoing).

• - Phonologische Parameter (z.B. betreffend Phoneme, Silben);
• - Morphologische Parameter (z.B. betreffend Bestandteile einer Sprache (Morpheme), deren Kombination, Verhältnis zueinander, etc.);
• - Syntaktische Parameter (z.B. zur Form und der Struktur von Wortkombinationen, Satzgrammatik);
• - Lexikologische Parameter (z.B. betreffend einlernen bestimmter Wörter);
• - Semantische Parameter (z.B. die Bedeutung von Wörtern in Bezug auf bestimmte Gegebenheiten);
• - Satzsemantische Parameter (z.B. betreffend logische Bedeutung von Sätzen);
• - Textsemantische Parameter (z.B. betreffend Bezüge auf erkannte Gegebenheiten).

In this example, the following parameters of the linguistic communication between the occupants can be detected, evaluated by means of speech recognition and taken into account:

• - Phonological parameters (eg concerning phonemes, syllables);
• - Morphological parameters (eg concerning components of a language (morphemes), their combination, relationship to each other, etc.);
• - Syntactic parameters (eg on the form and structure of word combinations, sentence grammar);
• - lexicological parameters (eg regarding learning specific words);
• - semantic parameters (eg the meaning of words in relation to specific circumstances);
• - sentence-semantic parameters (eg regarding the logical meaning of sentences);
• - Text-semantic parameters (eg regarding references to recognized circumstances).

Furthermore, voices of one or more inmates can be learned in the method and taken into account, in particular in connection with content-related features. Teach-in voices can be represented as one or more typical parameter values for the voice. On the basis of such parameters, identification data for one or more specific voices can be determined or processed in the method. For example, the recognition data for one or more voices may also be obtained based on the data of a voice control, hands-free device of the vehicle, or a mobile user device (e.g., when making a phone call).

One or more of the mentioned parameters are also varied during the adaptation of the at least one auditory message to be output, in particular by means of a speech generation or speech synthesis, in particular such that an optimization according to predetermined criteria is sought. In this case, an appropriate combination of the parameters can be determined and taken into account.

In this example, the device is designed such that a mutual interference between an auditory message and a communication between at least two occupants, in particular with respect to the audibility and / or intelligibility of linguistic contents, is minimized. With the method, this can also be achieved with an optimal volume of speech communication and / or auditory communication.

An auditory message can be adjusted in the method such that a piece of information represented by tones, sequences of tones or melodies is varied. Alternatively or additionally, in turn, tones, sequences of tones or melodies can be adapted, which represent a substantially unchanged information.

In other words, information to be transmitted with the auditory message, for example based on vehicle data, e.g. a driver information can be varied. This can be output in the method by means of modified, adapted tones, sequences of tones or melodies. In this case, at least a first criterion of an auditory message with a certain type or a specific content can be adapted, while at least a second criterion, which identifies the auditory message as an auditory message of the particular kind or with the particular content, is issued or enhanced unchanged. For example, one or more frequencies and / or amplitudes in which the tones of the auditory message are output may be adjusted, while a sequence of the tones, for example only the temporal relationships between the output of the tones, remain substantially constant. Thus, it can be achieved that the auditory message is still sufficiently recognizable or interpretable by human perception.

In another example, one or more priority measures for the present communication can be determined from a content characteristic of the linguistic communication, in particular taking into account (matching) with at least one content-related parameter from the history, and taken into account in the method.

It can be stated in a simplified example that it is an important conversation of the inmates, and then generating the auditory message are suppressed and made up for later.

Depending on the alignment of the read-in, determined or estimated priority measures, it is possible to adapt at least one auditory message to be output. In this case, the time of the output and / or at least one, in particular acoustic parameters in the output of the auditory message can be varied. For example, if only a short time interval between speech communication is detected or predicted, an auditory message may be issued at the appropriate time interval. In this case, at least the time or a time interval of the output of the auditory message is adjusted. Further, the method may also adjust the time at which the auditory message is to be ended and / or the auditory message time interval. In particular, the auditory message may be appropriately placed or scheduled within an available time interval within a communication between the at least two occupants of the vehicle.

For example, this may be output between two phrases of a speaker or a phrase of a speaker, wherein a time interval until an expected response of the phrase by another occupant is sufficiently determined, predicted or estimated. Further, depending on the matching of said parameters, information intended to be output as an audible message may alternatively or additionally be presented as a visual message, e.g. Image, graphic, text or symbol, or a haptic message, e.g. Vibration, be issued.

For example, if it is determined from the adjustment of the parameter that the acoustic content is to receive an overall higher priority than the acoustic message, the auditory message can be suppressed or output at a later time and / or with changed parameters. For example, if it is determined from the alignment of the priority measures that a particular auditory message has an overall higher priority measure than a certain auditory content, the auditory message may be associated with a changed parameter, e.g. louder or predominantly from another loudspeaker, and in addition a visual or graphical and / or haptic message are output. One or more auditory messages may be generated in the method on this basis, wherein the times for their output are chosen such that, with respect to the output time interval, they are substantially separated between two further audible contents, e.g. to certain thresholds exceeding amplitudes or amplitude maxima of two acoustic content, fits.

To output the auditory message, an emotion recognition and / or intention recognition of an occupant of the vehicle can be carried out, after which an adaptation of the type and / or volume and / or intonation and / or pitch of the auditory message takes place. The emotion recognition and / or intent detection can with the help of sensors, such as the camera 6 , the vehicle can be detected. For this purpose, an evaluation of the time course of the facial expressions of the occupants and / or their gestures and / or other physical impulses can take place.

With a recognized, high emotional excitement of at least one of the inmates, in particular of the speaking inmate, an earlier and / or stronger adjustment of the auditory message can take place. This can increase the visibility of the auditory message and the traffic safety.

In another example of the method, an auditory message may be varied in terms of its verbal, tone, voice, linguistic parameters, depending on the content parameter of the linguistic communication between at least two occupants of the vehicle, taking into account a so-called communication model. In this case, a specific, for example dependent on the user settings, communication goal can be sought.

• - einem psychischen Bedürfnis zumindest einer Person im Wohnraum entsprochen wird, und/oder
• - ein Informationsaustausch verbessert, insbesondere optimiert, wird, und/oder
• - ein bestimmtes Kommunikationsziel gefördert wird, und/oder
• - ein emotionaler Ausgleich gefördert wird.

For example, one or more auditory messages may be adjusted taking into account a communication model, preferably based on a transaction analysis or an SR model (= stimulus-response model). For example, the adjustment can be made such that

• - a mental need of at least one person in the living space is met, and / or
• an information exchange is improved, in particular optimized, and / or
• - a specific communication objective is promoted, and / or
• - an emotional balance is promoted.

The aspects of the method and apparatus described herein which have been applied to an auditory message may alternatively be applied to the adaptation of further audible content to be output in the vehicle, for example to information for at least one occupant, e.g. tones representative of driver information, sequences of tones or melodies, in a corresponding or adapted manner.

LIST OF REFERENCE NUMBERS

1

vehicle

2L

left front seat

2R

right front seat

3L

left, rear seat

3R

right, rear seat

4

multimedia system

5

On-board computer (arithmetic unit)

6

camera

7

microphone

8-1, ..., 8-7

speaker

9

Sensor technology (seat allocation, seat belt detection)

11, 12, 13

inmate

14

steering wheel

Claims (18)

A method of generating auditory messages from a computer carried by a vehicle (1), the auditory message comprising a voice message to occupants (11, 12, 13) of the vehicle (1), comprising the steps of: Determining at least one content-related feature of a linguistic communication between the occupants (11, 12, 13) of the vehicle (1), and - Adjust the auditory message depending on the determined content characteristic of the linguistic communication between the occupants (11, 12, 13) of the vehicle (1). Method according to Claim 1 in which the at least one content-related feature of the linguistic communication is determined by means of a linguistic analysis, wherein the linguistic analysis processes one or more of the following features of the linguistic communication: a semantics; - a theme feature; an information as to whether or not the linguistic communication comprises a content-concluded message, in particular sentences and / or phrases. Method according to Claim 1 or 2 , in which the content of the auditory message is adjusted depending on the determined content characteristic of the linguistic communication between the inmates. Method according to one of the preceding claims, in which a time for the output of at least part of the auditory message is determined as a function of the determined content-related feature. Method according to Claim 4 in which the auditory message is paused and / or divided into at least two parts as a function of the determined content-related feature. Method according to one of the preceding claims, in which the auditory message is adapted stylistically, with regard to its voice and / or pitch and / or undertone, depending on the determined content-related feature. Method according to one of the preceding claims, in which recognition of one or more keywords in the communication of at least one of the occupants (11, 12, 13) and / or recognition of the gestures of at least one of the occupants (11, 12, 13) is performed, and then the auditory message is adjusted in terms of their content feature and / or an acoustic parameter. Method according to one of the preceding claims, in which by means of a linguistic evaluation of the linguistic communication, a prediction of a temporal parameter, in particular of a chronological progression, of the linguistic communication is carried out and the temporal parameter is processed during the adaptation of the auditory message to be output. Method according to one of the preceding claims, in which the type and / or content of another audible context audible in the interior of the vehicle (1) is determined and processed as a function of an analysis result for adapting the generation and / or the output of the auditory message. Method according to one of the preceding claims, in which, depending on an evaluation of at least one physical posture and / or a gesture of the at least one occupant (11, 12, 13) of the vehicle (1), it is recognized that a linguistic message: explicitly to the computer carried in the vehicle (1) or explicitly not to the computer (5) of the vehicle (1), and / or - is directed to a specific occupant (11, 12, 13) of the vehicle (1). Method according to Claim 10 in which a probability measure for a recognized linguistic message is determined, wherein if the probability measure exceeds a threshold that the message is not addressed to the computer (5), the auditory message is adjusted. Method according to one of the preceding claims, in which at least one feature of the linguistic communication between the at least two occupants (11, 12, 13) of the vehicle (1) depends on the recognized body language and / or a voice position, in particular a change of the voice, at least one of the occupants (11, 12, 13) is determined and processed to adapt the auditory message. Method according to one of the preceding claims, in which a sequence of at least two auditory messages and / or individual contents of a respective auditory message depends on a current communication or a predicted communication project and / or a voice of at least one occupant (11, 12, 13) and / or an occupant (11, 12, 13) at a particular seat is changed. Method according to one of the preceding claims, in which a content-related feature of the linguistic communication between at least two occupants (11, 12, 13) of the vehicle (1) is determined, and / or - the nature and / or priority of one or more auditory reports to be issued is determined; and then: a sequence of the at least two auditory reports and / or individual contents of the auditory report is determined, depending on the determined content-related feature of the spoken communication and / or the ascertained type and / or priority of the at least two auditory messages to be output. Method according to one of the preceding claims, in which - it is determined that a part of the auditory message is not sufficiently perceptible with a probability that exceeds a predetermined threshold, - at least part of the auditory report associated with the imperceptible part, in particular the part of the auditory report, is repeated. A computer program product that can be loaded directly into the internal memory of a digital computer and includes software code portions that perform the steps of any one of the preceding claims when the product is run on a computer. Apparatus for generating auditory messages from a computer carried by a vehicle (1), the auditory message comprising a voice message to occupants (11, 12, 13) of the vehicle (1), comprising: a first means for determining at least one content-related feature of a linguistic communication between the occupants (11, 12, 13) of the vehicle (1), and - A second means for adjusting the auditory message depending on the determined content characteristic of the linguistic communication between the occupants (11, 12, 13) of the vehicle (1). Device after Claim 17 , characterized in that this further means for performing the method according to one of Claims 2 to 15 includes.

Images