Method and system for configuring a display representation for a display

Abstract

The invention relates to a method for configuring a display representation for a display (50), wherein at least one user input (10) is detected by means of at least one input device (2), wherein, based on the detected at least one user input (2), a configuration file (20) for the display representation, formulated in a description language, is generated by means of a generative model (5), and wherein a display signal (30) for the display (50) is generated and provided based on the configuration file (20) formulated in the description language. The invention further relates to a system (1) for configuring a display representation for a display (50).

Description

[0001] The invention relates to a method and a system for configuring a display representation for a display.

[0002] With increasing digitalization in vehicles, more and more display elements and information in vehicles are presented in purely digital form on displays. This enables flexible configuration of the display presentation, i.e. the way in which information is arranged and presented on the display. However, with an increasing amount of information (vehicle data, navigation map, infotainment data, etc.), the complexity of the configuration also increases. With existing displays, it is possible to choose between various predefined views. It can be assumed that the customizability of displays in vehicles will continue to increase. For example, infotainment displays already have a "home screen" on which various information displays (navigation map, fuel consumption, album cover, etc.) as well as abbreviations for vehicle functions (e.g.Telephone speed dial, media control, jumping to vehicle settings, etc.) can be individually arranged using a grid.

[0003] An increasing number of options generally also leads to more complex operation, which can increasingly overwhelm users. While it is theoretically possible to configure and customize the display (almost) completely freely, this may be used less frequently because the complexity prevents users from configuring the display as desired. Furthermore, it is difficult to implement more complex modeling, such as changing the display when engaging reverse gear, with the input devices available in the vehicle (touch controls, buttons). Consequently, a simple yet powerful user interface is required to utilize the full customization potential.

[0004] Methods and systems for configuring display representations are known from EP 3 715 164 A1, GB 2 592 217 A, US 2021 / 0 237 572 A1, US 2020 / 0 047 617 A1, US 2022 / 0 289 029 A1.

[0005] The invention is based on the object of improving a method and a system for configuring a display representation for a display, in particular with regard to a configuration of the display representation.

[0006] The object is achieved according to the invention by a method having the features of patent claim 1 and a system having the features of patent claim 10. Advantageous embodiments of the invention emerge from the subclaims.

[0007] In particular, a method for configuring a display representation for a display, in particular for a display in a vehicle, is provided, wherein at least one user input is detected by means of at least one input device, wherein, based on the detected at least one user input, a configuration file for the display representation formulated in a description language is generated by means of a generative model, and wherein a display signal for the display is generated and provided based on the configuration file formulated in the description language.

[0008] Furthermore, in particular, a system is created for configuring a display representation for a display, in particular for a display in a vehicle, comprising at least one input device which is configured to detect at least one user input, a data processing device, wherein the data processing device is configured to generate a configuration file for the display representation formulated in a description language based on the detected at least one user input by means of a generative model, and a signal generation device, wherein the signal generation device is configured to generate and provide a display signal for the display based on the configuration file formulated in the description language.

[0009] The method and the system enable a display to be configured in a simple manner. For this purpose, at least one user input is captured by at least one capture device. The user input relates to the display, i.e., a user describes the desired display using the user input. Based on the at least one user input, a generative model is used to generate a configuration file for the display formulated in a description language. In other words, the generative model is trained in particular to generate the configuration file (in particular as a text file) based on the at least one user input. Based on the configuration file formulated in the description language, a display signal for a display is generated and provided. This is done by means of a signal generation device.In particular, it is intended that the generated display signal be output on a display. This allows the user to see the effect of the at least one user input. The user can then further adjust and configure the display representation, in particular by making additional user inputs.

[0010] The display is in particular a display in a vehicle. The display can, for example, be a combination display or an infotainment display in a vehicle. The method is in particular carried out at least partially in a vehicle. In particular, the detection of the at least one user input takes place in the vehicle. The display representation is in particular a display representation in the context of a vehicle. Elements and / or information of the display representation are in particular in an automotive context. Elements and / or information of the vehicle can, for example, comprise virtual instruments of the vehicle, vehicle data, navigation maps and / or navigation data and / or vehicle-specific infotainment data. The vehicle in particular comprises the at least one detection device. In one alternative, the vehicle in particular comprises the data processing device. The vehicle in particular comprises the signal generating device.The vehicle includes, in particular, the display on which the generated display signal is output. In particular, the system is arranged at least partially or entirely within the vehicle. A vehicle is, in particular, a motor vehicle. However, the vehicle can also be another land, rail, water, air, or space vehicle.

[0011] A generative model is, in particular, a model from the field of artificial intelligence and / or machine learning. The generative model is, in particular, a deep generative model. In particular, the generative model can be designed according to a transformer architecture (cf. Polosukhin, Illia; Kaiser, Lukasz; Gomez, Aidan N.; Jones, Llion; Uszkoreit, Jakob; Parmar, Niki; Shazeer, Noam; Vaswani, Ashish, June 12, 2017, Attention Is All You Need, arXiv:1706.03762 [cs.CL]). In particular, the generative model can be a generative pre-trained transformer model, e.g., GPT-3 or GPT-4, or be based on one. In particular, the generative model can be a generative language model. In particular, the generative model can be a large language model. The generative model can be based, for example, on StarCoder, Ilama-2, or Mistral-8x7b (Mistral.Al, France) (see below).

[0012] The description language can be, for example, one of the following: XML, JSON, HTML or any other suitable description language for a display representation.

[0013] For clarification, an example in an XML-based description language is given below.

[0014] For example, the following user input occurs: "I would like to configure the display for the cluster display. Please add a tube display on the left, but it should be slightly spaced from the left edge and from the top. This tube should display the vehicle speed. Additionally, when shifting into reverse gear, the image from the rearview camera should be inserted into the tube. However, as soon as another gear is selected, the vehicle speed should be displayed again."

[0015] The generative model then produces, for example, the following configuration file as output: (adapted from: https: / / www-archive.mozilla.org / docs / ora-oss2000 / chatzilla / xuldoc-example.html) <?xml version="1.0"?><!DOCTYPE window><?xul-overlay href="file: / / / u / rginda / src / HTML / moztests / sampleoverlay.xul"?> <window orient="vertical" xmlns="http: / / www.mozilla.org / keymaster / gatekeeper / there.is.only.xul"> <type>dcp < / type> <tube orient="left" style="classical"> <data> vehicle_speed < / data> <position> <unit> px < / unit> <value> 50 30 0 0< / value> < / position> <event type="onChange" trigger="gearselection"> <if value="rear"> <data> video:rear_camera < / data> < / if> <else> <data> default < / data> < / else> < / event> < / tube> < / window>

[0016] The generative model can either be fully trained for the process and system, or it can use a previously trained generative model that is then adapted to the process and system (fine-tuning). The process for a generative language model is as follows: in a first step, the language model is trained using a large amount of text, which can also be program code. In this case, the program code is the descriptive language for the display. This so-called pre-training serves in particular to build a kind of vocabulary and corresponding semantic relationships, or to learn them through the generative model. The result of this first step is a pre-trained basic model (a so-called foundation model). The foundation model can then be adapted / optimized for specific use cases through what is known as fine-tuning.For this purpose, pairs of at least one user input (i.e., a "description" by the user) and program code of the description language are presented as examples (using supervised learning). The advantage is that relatively few examples are required, since the generative model has already built up basic knowledge through pre-training and is therefore familiar with many connections. Fine-tuning therefore primarily serves to teach the generative model, especially the generative language model, what certain instructions mean and what the corresponding result is in the description language (or, more specifically, in the program code of the description language).

[0017] It may be intended to use an open source language model, for example, a Foundation Model or a model that has been trained to generate program code (such as StarCoder, see Raymond Li et al., StarCoder: may the source be with you!, Computation and Language (cs.CL), arXiv:2305.06161 [cs.CL], https: / / doi.org / 10.48550 / arXiv.2305.06161; or Llama-2, https: / / ai.meta.com / llama / from Meta AI). By providing examples consisting of instructions for the desired display representation and how this would look in the program code of the description language, the generative model can then be explicitly adapted to the procedural use case (fine-tuning). For example, the generative model can be trained for specific graphical elements ("widgets") and / or event mechanisms by means of fine-tuning.In this way, the generative model can be trained precisely to ensure that the intended graphical elements and event mechanisms can be integrated.

[0018] Parts of the system, in particular the data processing device and the signal generating device, can be implemented individually or collectively as a combination of hardware and software, for example as program code executed on a microcontroller or microprocessor. However, it can also be provided that parts are implemented individually or collectively as an application-specific integrated circuit (ASIC) and / or a field-programmable gate array (FPGA). Furthermore, in particular, graphical processing units (GPUs) and / or tensor processing units (TPUs) and / or M-processors with a neural engine can also be used to provide the generative model.

[0019] In one embodiment, the user input is recorded in dialog form. This allows the configuration of the display to be carried out step by step. In other words, the user can make a user input, based on which the configuration file is generated, and the display is shown on a display according to the configuration file. The system can then ask whether a change or adjustment to the display is desired. If this is the case, the user makes another user input, and the configuration file is regenerated and / or adjusted using the generative model. The thus modified and / or supplemented display is shown on the display, etc. If, on the other hand, the user is satisfied with the display, the configuration can be ended. The system can ask questions and / or provide instructions, etc.for example, display it in text form on a display and / or output it via a voice output.

[0020] In one embodiment, the configuration file is created in a human-readable description language. This allows the configuration file to be subsequently manually adjusted if further changes are desired. For example, XML, JSON, or HTML can be used for this purpose. A human user with the appropriate knowledge can directly understand the human-readable description language and make changes to the configuration file.

[0021] In one embodiment, it is provided that the at least one user input comprises at least one voice input, wherein the generative model is designed as a generative language model, and wherein the at least one voice input is fed to the generative language model. This allows a user to configure the display representation by means of spoken instructions. In particular, it is provided that the voice input is formulated in the form of instructions and / or in the form of a dialogue between the user and the system (e.g. the voice input can include the following content: “Please generate a combination display representation with a tube on the left, a map representation on the right and a visualization of the environmental data in the middle”). During training, in particular during fine-tuning, example pairs of voice inputs (orText inputs) paired with the corresponding program code in the description language.

[0022] In one embodiment, the at least one speech input is converted into text using speech recognition, with the converted text being fed to the generative language model. This allows the generative language model to be used more versatilely, as direct text input is also possible. Furthermore, pre-trained generative language models can be used, such as GPT-3 or GPT-4, StarCoder, or Llama-2. During training, particularly during fine-tuning, the generative model, in particular the generative language model, is presented with example pairs of text inputs paired with the respective associated program code in the description language.

[0023] In one embodiment, the at least one user input comprises at least one gesture detected by a touch-sensitive operating device, wherein the detected at least one gesture is fed to the generative model. This allows a user to generate a desired display representation, for example using gestures. For example, symbols (window icons, abbreviations, etc.) can be sketched onto the touch-sensitive operating device using an actuating element (touch pen or finger), from which the configuration file is then generated. During training, in particular during fine-tuning, example pairs of gestures or traces of the gestures are presented to the generative model, paired with the respective associated program code in the description language.

[0024] In one embodiment, it is provided that the at least one user input comprises at least one text input, wherein the generative model is designed as a generative language model, and wherein the at least one text input is fed to the generative language model. In this way, the display representation can be configured via a text input. In particular, it is provided that the text input is formulated in the form of prose text and / or in the form of a dialogue (e.g., the following text can be entered: “Please generate a combination display representation with a tube on the left, a map representation on the right, and a visualization of the environmental data in the middle”). The text input can be made, in particular, using a virtual or physical keyboard as the input device.During training, especially during fine-tuning, the generative model, in particular the generative language model, is presented with example pairs of text inputs paired with the corresponding program code in the description language.

[0025] It can be provided that different types of at least one user input are combined. This can increase flexibility during configuration. For example, part of the configuration can be performed using at least one voice or text input, while another part can be performed using at least one gesture.

[0026] In one embodiment, the generative model is provided and executed locally in a vehicle. This allows a display representation to be configured even without a connection to a backend.

[0027] In one embodiment, the at least one user input is transmitted to a backend server, with the generative model being provided and executed on the backend server. This allows the generative model to be provided and maintained centrally. Furthermore, the use of a backend server allows for the use of more powerful generative models, since more computing and storage resources can be provided and utilized than with a local application in a vehicle.

[0028] In one embodiment, the generative model additionally considers a current configuration file and / or previous user inputs. This allows a current display representation and / or previous user inputs to be taken into account, so that the generative model can relate further, subsequent user inputs to the current display representation and the previous user inputs. This enables a step-by-step configuration of the display representation in a common context.

[0029] In one embodiment, the configuration file is post-processed on a data processing device after creation. This allows the display representation to be edited in detail in the configuration file itself. In particular, users with appropriate knowledge can configure the display representation directly in the configuration file or adapt the content of a configuration file created according to the method.

[0030] In one embodiment, a configuration file generated according to the method is used together with the associated at least one user input to train the generative model. This allows the generative model to be continuously trained.

[0031] In one embodiment, a configuration file generated according to the method is stored on a server and can be retrieved from there for use. This allows generated configuration files to be used for display representations by multiple users and / or shared among them.

[0032] In one embodiment, generated display representations and / or configuration files are identified and stored using versioning. These display representations and / or configuration files can be retrieved at a later time and used as the active display representation and / or configuration file. This makes it easy to return to a previous version in the event of errors and / or unwanted changes. The versions of the configuration files can be stored locally in a vehicle and / or on a backend server.

[0033] Further features of the system's design are described in the various embodiments of the method. The advantages of the system are the same as those of the various embodiments of the method.

[0034] The invention will be explained in more detail below using preferred embodiments with reference to the figures. Fig. 1 is a schematic representation of an embodiment of the system for configuring a display representation for a display; Fig. 2 a schematic flow diagram to illustrate an embodiment of the method for configuring a display representation for a display.

[0035] The Fig. 1 shows a schematic representation of an embodiment of the system 1 for configuring a display representation for a display 50, in particular in a vehicle. The system 1 comprises at least one input device 2, a data processing device 3, and a signal generating device 4. The system 1 is particularly configured to carry out the method described in this disclosure. The method is explained in more detail below using the system 1.

[0036] The input device 2 is configured to capture at least one user input 10.

[0037] The data processing device 3 comprises, in particular, a computing device 3-1 and a storage device 3-2. The data processing device 3 is configured to generate, based on the detected at least one user input, a configuration file 20 for the display representation, formulated in a description language, by means of a generative model 5.

[0038] The generative model 5 is stored in the storage device 3-2, in particular in the form of a structural description and associated parameters (e.g. weights, etc.).

[0039] The signal generating device 4 is configured to generate and provide a display signal 30 for the display 50 based on the configuration file 20 formulated in the description language. The display signal 30 contains the display representation according to the configuration file 20.

[0040] It may be provided that the user input 10 is captured in dialog form. In particular, it may be provided that the system 1 is configured to ask a user questions and / or instructions in dialog form, for example as a voice output and / or text display.

[0041] It may be provided that the configuration file 20 is generated in a human-readable description language. In particular, it may be provided that the configuration file is generated in one of the following description languages: XML, JSON, or HTML, etc.

[0042] It can be provided that the at least one user input 10 comprises at least one speech input 11, wherein the generative model 5 is designed as a generative speech model, and wherein the at least one speech input 11 is fed to the generative speech model. For this purpose, the at least one detection device 2 comprises, in particular, a microphone 2-1.

[0043] In a further development, it can be provided that the at least one speech input 11 is converted into text by means of speech recognition, wherein the converted text is fed to the generative language model 5. For this purpose, the capture device 2 can have a speech-to-text conversion module (not shown). However, the speech-to-text conversion can also be performed in the data processing device 3 or another device separate therefrom (not shown).

[0044] It can be provided that the at least one user input 10 comprises at least one gesture 12 detected by means of a touch-sensitive operating device 2-2, wherein the detected at least one gesture 12 is fed to the generative model 5. The at least one detection device 2 comprises the touch-sensitive operating device 2-2 and / or is designed as such.

[0045] It can be provided that the at least one user input 10 comprises at least one text input 13, wherein the generative model 5 is designed as a generative language model, and wherein the at least one text input 13 is fed to the generative language model. For this purpose, the at least one capture device 2 comprises, in particular, a virtual or physical keyboard 2-3.

[0046] It may be provided that the generative model 5 is provided and executed locally in a vehicle. The vehicle is, in particular, a motor vehicle. However, the vehicle can also be another land, rail, water, air, or space vehicle.

[0047] It can be provided that the at least one user input 10 is transmitted to a backend server 6, wherein the generative model 5 is provided and executed on the backend server 6. The generated configuration file 20 can then be transmitted back to the location of the application. If, for example, a display representation of a display 50 is configured in a vehicle, the at least one user input 10 is also captured there by means of the at least one capture device 2. The at least one user input 10 is then transmitted from the vehicle to the backend server 6. There, the configuration file 20 is generated based there and transmitted back to the vehicle. In the vehicle, the signal generation device 4 arranged there generates the display signal 30 based on the configuration file 20, which is output on the display 50.Further user inputs 10 can then be made to further configure the display representation, whereby the process is repeated.

[0048] It can be provided that the generative model 5 additionally takes into account a current configuration file 20 and / or previous user inputs 10. In particular, the current configuration file 20 and / or the previous user inputs 10 are supplied to the generative model 5 for this purpose, or the generative model 5 already takes these into account structurally, i.e., user inputs 10 received by the generative model 5 and configuration files 20 generated by the generative model 5 are already contained in a memory of the generative model 5.

[0049] It may be provided that the configuration file 20 is post-processed on a data processing device 60 after it has been created. For example, a user can adapt the configuration file 20 on a desktop computer, laptop, tablet computer, or smartphone in order to change and / or correct individual details.

[0050] It can be provided that a configuration file 20 generated according to the method is used together with the associated user input 10 to train the generative model 5. This takes place in particular on a backend server 6. In particular, a plurality of generated configuration files 20 with the associated user inputs 10 can be used for training.

[0051] It can be provided that a configuration file 20 generated according to the method is stored on a server 70 and can be retrieved from there for use, for example, by the data processing device 3. The server 70 can also be the backend server 6. The server 70 can, for example, provide a platform on which configuration files 20 can be uploaded, searched, and retrieved. A retrieved or downloaded configuration file 20 can then be edited and adapted again using the system 1 and the method.

[0052] It can be provided that generated display representations and / or configuration files 20 are identified and stored using versioning. These display representations and / or configuration files 20 can be retrieved at a later time and used as the active display representation and / or configuration file 20. This makes it easy to return to a previous version in the event of errors and / or unwanted changes. The versions of the configuration files 20 can be stored locally in a vehicle and / or on the backend server 6.

[0053] The Fig. 2 shows a schematic flow diagram to illustrate an embodiment of the method for configuring a display representation for a display, in particular in a vehicle.

[0054] In measure 100, at least one user input is recorded using at least one input device.

[0055] In measure 101, based on the recorded at least one user input, a configuration file for the display representation formulated in a description language is generated using a generative model.

[0056] In measure 102, a display signal for the display is generated and provided from the configuration file formulated in the description language.

[0057] In a measure 103, the display signal is displayed on the display so that a user can capture the display representation and check the configuration made.

[0058] You can then return to step 100 to continue with the configuration and / or make further adjustments to the display based on the configuration file created in step 101.

[0059] The following describes an example process flow for configuring the display of a display in a vehicle. In particular, it is envisaged that user inputs are entered or recorded in the form of a dialogue with the system. For example, it is assumed that the user inputs are voice inputs that are converted into text and fed to the generative model as text inputs. Alternatively, text inputs can also be used. In principle, however, direct voice input, a gesture, or other user inputs are also possible. It can also be envisaged to combine different types of user input. For example, part of the configuration can be carried out using at least one voice or text input, while another part can be carried out using at least one gesture.

[0060] In a first cycle, the following voice input from a user is recorded in measure 100: “Please create a combined display with a tube on the left, a map on the right, and a visualization of the surrounding data in the middle.”

[0061] In step 101, the generative model, specifically the generative language model, generates the configuration file formulated in the description language based on this user input. In step 102, the display signal is generated from this file and displayed on the display in step 103. The process then returns to step 100.

[0062] In a second cycle, the following voice input is recorded in measure 100: "Okay, that looks good. Please enlarge the tube on the left, about 10%. Please display the map in a rectangle with rounded corners."

[0063] In step 101, the generative model, specifically the generative language model, generates an adapted configuration file based on this user input. The previous display representation or the previous configuration file and the previous user input are considered as context. In step 102, the display signal is generated from this and displayed on the display in step 103. The process then returns to step 100.

[0064] In a third cycle, the following voice input is recorded in measure 100: "For the visualization of the surrounding data, can you please set a rule so that when it gets dark, the image from the night vision device is automatically displayed? However, when I select reverse gear, Area View should be visible."

[0065] In step 101, the generative model, specifically the generative language model, generates an adapted configuration file based on this user input. The previous display representations or the previous configuration files and the previous user inputs are considered as context. In step 102, the display signal is generated from this and displayed on the display in step 103. The process then returns to step 100.

[0066] The generative model, especially the generative language model, provides feedback to the user that the rules have been stored. The system can also query the user to confirm whether the configuration is complete.

[0067] In a fourth cycle, the following speech input is recorded in measure 100: „ Not quite yet. Could you please change the background color to a vibrant red with a hint of orange?"

[0068] In step 101, the generative model, specifically the generative language model, generates an adapted configuration file based on this user input. The previous display representation or the previous configuration file is considered as context. In step 102, the display signal is generated from this and displayed on the display in step 103. The process then returns to step 100.

[0069] Subsequently, further cycles of the procedure can be carried out. List of reference symbols 1 system 2 Input device 2-1 Microphone 2-2 touch-sensitive control device 2-3 Keyboard 3 Data processing facility 3-1 Calculation device 3-2 Storage device 4 Signal generating device 5 generative model 6 backend servers 10 User input 11 Voice input 12 gesture 13 Text input 20 Configuration file 30 display signal 50 ad 60 Data processing facility 70 servers 100-103 Measures of the procedure QUOTES CONTAINED IN THE DESCRIPTION

[0000] This list of documents submitted by the applicant was generated automatically and is included solely for the convenience of the reader. This list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions. Cited patent literature

[0000] EP 3 715 164 A1

[0004] GB 2 592 217 A

[0004] US 2021 / 0 237 572 A1

[0004] US 2020 / 0 047 617 A1

[0004] US 2022 / 0 289 029 A1

[0004] Cited non-patent literature

[0000] Polosukhin, Illia; Kaiser, Lukasz; Gomez, Aidan N.; Jones, Llion; Uszkoreit, Jacob; Parmar, Nick; Shazeer, Noam; Vaswani, Ashish, June 12, 2017, arXiv:1706.03762

[0011] https: / / www.archive.mozilla.org / docs / ora-oss2000 / chatzilla / xuldoc-example.html) https: / / doi.org / 10.48550 / arXiv.2305.06161

[0017] Raymond Li et al., Computation and Language (CS.CL), arXiv:2305.06161 [CS.CL], https: / / doi.org / 10.48550 / arXiv.

Claims

[1] Method for configuring a display representation for a display (50), wherein at least one user input (10) is detected by means of at least one input device (2), wherein, starting from the detected at least one user input (2), a configuration file (20) formulated in a description language for the display representation is generated by means of a generative model (5), and wherein a display signal (30) for the display (50) is generated and provided from the configuration file (20) formulated in the description language. [2] Method according to claim 1, characterized by that the user input (2) is recorded in dialog form. [3] Method according to claim 1 or 2, characterized by that the configuration file (20) is created in a human-readable description language. [4] Method according to one of the preceding claims, characterized bythat the at least one user input (2) comprises at least one speech input (11), wherein the generative model (5) is designed as a generative language model, and wherein the at least one speech input (11) is supplied to the generative language model. [5] Method according to claim 4, characterized by that the at least one speech input (11) is converted into a text by means of speech recognition, wherein the converted text is fed to the generative language model. [6] Method according to one of the preceding claims, characterized by that the at least one user input (2) comprises at least one gesture (12) detected by means of a touch-sensitive operating device (2-2), wherein the detected at least one gesture (12) is fed to the generative model (5). [7] Method according to one of the preceding claims, characterized bythat the at least one user input (2) comprises at least one text input (13), wherein the generative model (5) is designed as a generative language model, and wherein the at least one text input (13) is fed to the generative language model. [8] Method according to one of the preceding claims, characterized by that the generative model (5) is deployed and executed locally in a vehicle. [9] Method according to one of claims 1 to 7, characterized by that the at least one user input (2) is transmitted to a backend server (6), wherein the generative model (5) is provided and executed on the backend server (6). [10] System (1) for configuring a display representation for a display (50), comprising: at least one input device (2) which is configured to detect at least one user input (10), a data processing device (3), wherein the data processing device (3) is configured to generate, based on the detected at least one user input (2), a configuration file (20) for the display representation formulated in a description language by means of a generative model (5), and a signal generating device (4), wherein the signal generating device (4) is configured to generate and provide a display signal (30) for the display (50) based on the configuration file (20) formulated in the description language.

Images