Vehicle machine tone intelligent recommendation method, electronic device, storage medium and program product

Abstract

The application provides a vehicle machine tone intelligent recommendation method, an electronic device, a storage medium and a program product, relates to the technical field of vehicle speech synthesis, and comprises the following steps: acquiring vehicle model identification information of a vehicle; performing analysis processing on the vehicle model identification information to obtain a vehicle model category of the vehicle; constructing a vehicle model portrait of the vehicle based on the vehicle model category; and determining at least one first target vehicle machine tone from a plurality of preset vehicle machine tones based on the cosine similarity between the vehicle model portrait and the plurality of preset vehicle machine tones to perform recommendation processing. The application can solve the problem that the traditional vehicle machine TTS tone selection depends on manual selection of a user and the operation process is complicated, thereby improving the accuracy of tone recommendation.

Description

Technical Field

[0001] This application relates to the field of vehicle speech synthesis (Text-to-Speech, TTS) technology, and in particular to a method for intelligent recommendation of vehicle infotainment system voice tone, electronic device, storage medium and program product. Background Technology

[0002] Currently, Text-to-Speech (TTS) technology has become a core foundational capability of intelligent vehicle infotainment systems, widely applied in various in-vehicle scenarios such as navigation announcements, voice assistant interactions, and multimedia voice broadcasts. In the application of TTS technology, the appropriate selection of voice timbre directly determines the auditory experience of human-vehicle voice interaction, and also reflects the brand style and model positioning of the automaker. Therefore, matching users with suitable voice timbres has become an important development direction for intelligent vehicle infotainment systems to optimize the interactive experience and create a unique style for each model. However, in related technologies, the voice timbre configuration schemes for vehicle infotainment systems all adopt a manual selection mode. This mode requires users to actively enter the multi-level settings interface of the vehicle infotainment system, manually browse and select the desired voice timbre. The overall operation path is cumbersome and the interaction steps are lengthy. This leads to most users never modifying the default voice timbre due to the complexity of the operation and a lack of willingness to actively configure it, thus using the system's initial default voice timbre for a long time. Furthermore, the fixed default voice timbre cannot take into account the auditory preferences, age characteristics, and usage habits of different users, nor can it adapt to the brand style of different levels and positioning of vehicle models, generally resulting in a serious mismatch between the voice timbre and user characteristics and vehicle model positioning. Summary of the Invention

[0003] The main objective of this application is to propose a method, electronic device, storage medium, and program product for intelligent recommendation of vehicle audio timbre, aiming to solve the problem that traditional vehicle audio TTS timbre selection relies on manual selection by the user, which is cumbersome and complicated, thereby improving the accuracy of audio timbre recommendation.

[0004] To achieve the above objectives, a first aspect of this application proposes a method for intelligent recommendation of vehicle infotainment system audio timbre, the method comprising: Obtain vehicle model identification information; The vehicle identification information is parsed to obtain the vehicle type category; Construct a vehicle model profile based on the vehicle model category; Based on the cosine similarity between the vehicle model profile and multiple preset vehicle infotainment system tones, at least one first target vehicle infotainment system tones is determined from the multiple preset vehicle infotainment system tones for recommendation processing.

[0005] In some embodiments, constructing the vehicle model profile based on the vehicle model category includes: The brand tone information of the vehicle is determined from a preset brand tone database based on the vehicle type category; Based on the brand tone information and the target user group information, price range information, and power type information corresponding to the vehicle type, a vehicle profile is constructed.

[0006] In some embodiments, the step of determining at least one first target vehicle infotainment sound from the plurality of preset vehicle infotainment sound sources based on the cosine similarity between the vehicle model profile and a plurality of preset vehicle infotainment sound sources for recommendation processing includes: The target cosine similarity between the vehicle model profile and multiple preset vehicle infotainment system timbres is determined from a preset vehicle model timbre mapping table; the vehicle model timbre mapping table includes: multiple vehicle model profiles, multiple preset vehicle infotainment system timbres, and the cosine similarity between the multiple vehicle model profiles and the multiple preset vehicle infotainment system timbres. At least one first target vehicle infotainment sound tone is determined based on multiple target cosine similarities for recommendation processing; among the multiple target cosine similarities, the cosine similarity between the first target vehicle infotainment sound tone and the vehicle model profile is higher than the cosine similarity between other vehicle infotainment sound tones and the vehicle model profile.

[0007] In some embodiments, after performing the recommendation process by determining at least one first target vehicle audio tone from the plurality of preset vehicle audio tones, the method further includes: Obtain implicit behavioral information about the user's response to the first target vehicle infotainment system's sound timbre in the vehicle; The explicit audio preferences and driving habits of the vehicle's users are combined with the implicit behavioral information and the time period features corresponding to the current time to form a dynamic user profile through multimodal fusion processing. Based on the dynamic user profile, the first target vehicle infotainment system's audio timbre is updated according to user preferences to obtain the updated audio timbre.

[0008] In some embodiments, the multimodal fusion processing of the user's explicit audio preferences and driving habit information in the vehicle with the implicit behavioral information and the time period features corresponding to the current time includes: Based on preset dynamic weights, the preference vectors of the vehicle infotainment system's explicit sound preference information, driving habit information, implicit behavioral information, and time period features are weighted and summed. The dynamic weights include explicit preference weights, driving habit weights, implicit behavior weights, and time period feature weights. The explicit preference weights are greater than the driving habit weights, implicit behavior weights, and time period feature weights. The driving habit weights, implicit behavior weights, and time period feature weights are dynamically adjusted based on the data volume of the driving habit information, the implicit behavior information, and the time period feature, respectively.

[0009] In some embodiments, the dynamic user profile includes dynamic user profiles of multiple users of the vehicle, and the method further includes: Obtain the user identification result of the vehicle; Based on the user identification results, a target user profile is determined from multiple dynamic user profiles; Based on the target user profile, the second target vehicle infotainment system audio is switched to the third target vehicle infotainment system audio; the second target vehicle infotainment system audio includes the first target vehicle infotainment system audio or the updated vehicle infotainment system audio, and the third target vehicle infotainment system audio is the vehicle infotainment system audio after the first target vehicle infotainment system audio is updated based on the target user profile by performing user preference updates.

[0010] In some embodiments, the method further includes: Obtain the usage scenario recognition results of the vehicle; Based on the usage scenario recognition result, the second target vehicle infotainment system audio timbre is subjected to parameter gradual adjustment processing to switch the second target vehicle infotainment system audio timbre to the fourth target vehicle infotainment system audio timbre; the second target vehicle infotainment system audio timbre includes the first target vehicle infotainment system audio timbre or the updated vehicle infotainment system audio timbre, and the fourth target vehicle infotainment system audio timbre is the vehicle infotainment system audio timbre corresponding to the usage scenario recognition result.

[0011] To achieve the above objectives, a second aspect of this application proposes an intelligent recommendation system for in-vehicle audio timbre, the system comprising: The acquisition module is used to acquire vehicle model identification information; The vehicle identification information parsing module is used to parse and process the vehicle identification information to obtain the vehicle type. The vehicle model profile building module is used to build a vehicle model profile based on the vehicle model category; The intelligent timbre recommendation module is used to determine at least one first target vehicle timbre from the multiple preset vehicle timbres based on the cosine similarity between the vehicle model profile and multiple preset vehicle timbres for recommendation processing.

[0012] To achieve the above objectives, a third aspect of this application provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the vehicle audio intelligent recommendation method described in the first aspect.

[0013] To achieve the above objectives, a fourth aspect of the present application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the vehicle infotainment system intelligent sound recommendation method described in the first aspect.

[0014] To achieve the above objectives, a fifth aspect of the present application provides a computer program product, which includes a computer program that, when executed by a processor, implements the vehicle infotainment system audio intelligent recommendation method provided in the first aspect above.

[0015] The intelligent recommendation method, system, electronic device, computer-readable storage medium, and computer program product for vehicle infotainment system tones proposed in this application involve: acquiring vehicle model identification information; parsing the vehicle model identification information to obtain the vehicle model category; constructing a vehicle model profile based on the vehicle model category; and determining at least one first target vehicle infotainment system tone from the multiple preset vehicle infotainment system tones for recommendation based on the cosine similarity between the vehicle model profile and multiple preset vehicle infotainment system tones.

[0016] In this embodiment, the vehicle model identification information is parsed to obtain the vehicle model category. Then, a vehicle model profile is constructed based on this model category. Based on the cosine similarity between this model profile and multiple preset vehicle infotainment system (VMS) voice prompts, at least one first target VMS voice prompt is determined from the multiple preset VMS voice prompts for recommendation. Thus, compared to the traditional solution that relies on cumbersome manual configuration of the VMS TTS voice prompt, this embodiment can achieve personalized TTS voice prompt recommendation with zero configuration by the vehicle user. Users can obtain a voice prompt that matches the vehicle model's tone without manual configuration, thereby solving the problem of cumbersome and complex manual selection in traditional VMS TTS voice prompts and improving the accuracy of voice prompt recommendations. Attached Figure Description

[0017] Figure 1 A flowchart illustrating the steps of the vehicle audio intelligent recommendation method provided in some embodiments of this application; Figure 2 for Figure 1 A detailed flowchart of step S103; Figure 3A flowchart illustrating the steps of the vehicle audio intelligent recommendation method provided in this application embodiment in other embodiments; Figure 4 The intelligent recommendation method for vehicle audio timbre provided in the embodiments of this application is illustrated in some embodiments with a schematic diagram of the vehicle identification number (VIN) parsing and vehicle model matching process. Figure 5 A flowchart illustrating the steps involved in the intelligent recommendation method for vehicle audio timbre provided in this application embodiment; Figure 6 The intelligent recommendation method for vehicle audio timbre provided in the embodiments of this application is illustrated in some embodiments with a flowchart of the user preference algorithm. Figure 7 A flowchart illustrating the steps of the vehicle infotainment system intelligent audio recommendation method provided in some other embodiments of this application; Figure 8 A flowchart illustrating the steps of the vehicle infotainment system intelligent audio recommendation method provided in some other embodiments of this application; Figure 9 The intelligent recommendation method for vehicle audio timbre provided in the embodiments of this application is illustrated in some embodiments with a schematic diagram of the scene recognition and audio timbre switching process. Figure 10 The system architecture diagram of the vehicle audio tone intelligent recommendation method provided in the embodiments of this application in some embodiments; Figure 11 This is a schematic diagram of the structure of the vehicle audio intelligent recommendation system provided in the embodiments of this application; Figure 12 This is a schematic diagram of the hardware structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0019] It should be noted that although functional modules are divided in the device / system schematic diagram and a logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the module division in the device / system or the order in the flowchart. The terms "first," "second," etc., in the specification, claims, and the aforementioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence.

[0020] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.

[0021] First, a brief explanation of the overall concept of the intelligent recommendation method for vehicle audio timbre provided in the embodiments of this application will be given.

[0022] In in-vehicle intelligent cockpits, TTS voice synthesis requires the selection of appropriate voice timbres for broadcasting. Voice timbre selection directly impacts user experience and brand image. For example, in terms of vehicle type, sporty cars suit youthful, energetic, and dynamic timbres; business cars suit calm, professional, and sophisticated timbres; family cars suit warm, friendly, and natural timbres; luxury cars suit high-end, elegant, and refined timbres; and new energy vehicles suit technologically advanced and futuristic timbres. In terms of user experience, gender preferences include male / female voices; style preferences include lively, calm, intellectual, and magnetic voices; speech speed preferences include fast, medium, and slow; and age-related preferences include: younger users prefer lively voices, while older users prefer calm voices. In terms of scenarios, navigation broadcasts need to be clear, concise, and at a moderate pace; entertainment interactions need to be lively, friendly, and emotional; vehicle control feedback needs to be brief, efficient, and clearly identifiable; and call modes need to be natural, fluent, and close to a human voice.

[0023] In related technologies, TTS (Text-to-Speech) voice selection relies on manual user settings and lacks intelligent recommendations. That is, traditional solutions require users to manually select voices in the settings interface, which is cumbersome. Furthermore, statistics show that over 80% of users have never changed the vehicle's default voice due to the complexity of the process and a lack of willingness to actively configure it, resulting in a voice mismatch between the user and the vehicle model. In addition, related technologies lack solutions for recommending vehicle voices during cold starts. That is, when a new user / vehicle uses the system for the first time, because no historical data is available, traditional solutions can only use a fixed default voice, failing to provide a personalized experience.

[0024] This means that traditional in-vehicle voice synthesis tone configuration schemes all adopt a manual selection mode. This mode requires users to actively enter the multi-level settings interface of the in-vehicle system, manually browse and select the desired broadcast tone. The overall operation path is cumbersome and the interaction steps are lengthy. As a result, most users never modify the default tone of the in-vehicle system due to the complexity of the operation and their lack of willingness to actively configure it, and thus use the initial default tone of the system for a long time. Furthermore, the fixed default tone cannot take into account the different auditory preferences, age characteristics, and usage habits of different users, nor can it adapt to the brand style of different levels and positioning of car models. There is a common problem that the broadcast tone is seriously mismatched with user characteristics and car model characteristics.

[0025] To address the aforementioned issues, this application proposes a method, system, electronic device, computer-readable storage medium, and computer program product for intelligent recommendation of vehicle-mounted audio tones. The aim is to solve the problem that traditional vehicle-mounted TTS audio selection relies on manual user selection, resulting in a cumbersome and complex process, thereby improving the accuracy of audio tone recommendations.

[0026] In this embodiment of the application, the vehicle model identification information is obtained; the vehicle model identification information is parsed to obtain the vehicle model category; a vehicle model profile is constructed based on the vehicle model category; and at least one first target vehicle model is determined from the multiple preset vehicle model sounds based on the cosine similarity between the vehicle model profile and multiple preset vehicle model sounds for recommendation processing.

[0027] In this embodiment, the vehicle model identification information is parsed to obtain the vehicle model category. Then, a vehicle model profile is constructed based on this model category. Based on the cosine similarity between this model profile and multiple preset vehicle infotainment system (VMS) voice prompts, at least one first target VMS voice prompt is determined from the multiple preset VMS voice prompts for recommendation. Thus, compared to the traditional solution that relies on cumbersome manual configuration of the VMS TTS voice prompt, this embodiment can achieve personalized TTS voice prompt recommendation with zero configuration by the vehicle user. Users can obtain a voice prompt that matches the vehicle model's tone without manual configuration, thereby solving the problem of cumbersome and complex manual selection in traditional VMS TTS voice prompts and improving the accuracy of voice prompt recommendations.

[0028] In some embodiments, a dynamic user profile can be constructed by fusing multimodal data such as explicit preferences, implicit behaviors, driving habits, and time-of-day characteristics based on a multimodal preference fusion algorithm. Based on this dynamic user profile, the user preference for the first target vehicle infotainment sound can be updated, thereby further improving the accuracy of sound recommendation.

[0029] In some embodiments, a scene-aware timbre smoothing mechanism can also be adopted. When automatically switching timbres based on scene recognition, parameter gradation is used to achieve a smooth transition of timbre switching, so that the user is unaware when the timbre changes with the scene, thereby improving the timbre transition naturalness score.

[0030] In some embodiments, the vehicle's current user can be automatically identified and the vehicle's audio settings can be switched according to the user's preferred audio tone.

[0031] Next, the intelligent recommendation method, system, electronic device, computer-readable storage medium, and computer program product for vehicle audio timbre provided in this application will be specifically described through the following embodiments, and the various detailed embodiments of the intelligent recommendation method for vehicle audio timbre provided in this application will be described in detail first.

[0032] It should be noted that in all specific embodiments of this application, when processing data related to user identity or characteristics, such as user information, user behavior data, user historical data, and user location information, user permission or consent is obtained first. Furthermore, the collection, use, and processing of this data comply with relevant laws, regulations, and standards. In addition, when embodiments of this application require access to sensitive personal information of users, separate permission or consent from the user is obtained through pop-ups or redirection to confirmation pages. Only after obtaining the user's separate permission or consent is the necessary user-related data required for the proper functioning of these embodiments acquired.

[0033] It should be noted that the intelligent vehicle audio recommendation method provided in this application relates to the field of vehicle voice synthesis technology. The intelligent vehicle audio recommendation method provided in this application can be applied to a terminal, a server, or software running on either a terminal or a server. In some embodiments, the terminal can be a vehicle, an in-vehicle terminal, or an electronic device such as a smartphone, tablet, laptop, or desktop computer associated with the vehicle and capable of communicating and interacting with it via a network. The server can be the terminal's backend server terminal device, which can be configured as an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms. The software can be an application implementing the intelligent vehicle audio recommendation method, a computer program, and a storage medium carrying the computer program. It should be understood that, based on different design needs in practical applications, the terminal, server, and software of the vehicle infotainment system with intelligent audio recommendation method provided in this application may also be other forms not listed here, and the vehicle infotainment system with intelligent audio recommendation method provided in this application does not specifically limit these.

[0034] Furthermore, this application can also be used in numerous general-purpose or special-purpose computer system environments or configurations. Examples include: vehicles, in-vehicle terminals, personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics devices, personal computers (PCs), minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices. This application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform specific tasks or implement specific abstract data types. This application can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via communication networks. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0035] For ease of understanding and explanation, the following text will use the vehicle-mounted terminal application of the intelligent vehicle-mounted audio recommendation method provided in the embodiments of this application as an example to describe the various specific embodiments of this application in detail. In some descriptions, the vehicle-mounted terminal may be simply referred to as an intelligent agent. The implementation of the intelligent vehicle-mounted audio recommendation method provided in the embodiments of this application for any of the above-mentioned subject matter can refer to the implementation process of the intelligent vehicle-mounted audio recommendation method described below.

[0036] Please refer to Figure 1 , Figure 1 The flowchart illustrates the steps of the vehicle infotainment system audio intelligent recommendation method provided in some embodiments of this application. It should be understood that, although... Figure 1 The flowcharts illustrating subsequent steps show the execution order of some method steps. However, based on different design needs in practical applications, the intelligent recommendation method for vehicle audio timbre provided in this application embodiment can, of course, employ a different execution order of method steps than shown in the figures. That is, Figure 1 The order of the method steps shown does not constitute a limitation on the execution logic order of the intelligent recommendation method for vehicle audio timbre provided in the embodiments of this application. Any other method based on this method is not limited to this method. Figure 1 Reasonable changes to the sequence of steps shown should be included within the protection scope of the intelligent vehicle audio recommendation method provided in the embodiments of this application.

[0037] like Figure 1 As shown, in some embodiments, the vehicle audio intelligent recommendation method provided in this application may include steps S101 to S104 as shown below.

[0038] Step S101: Obtain the vehicle model identification information.

[0039] In the cold start scenario of a new user / new car using the in-vehicle TTS, in order to achieve personalized tone recommendations with zero manual operation by the user, the vehicle terminal first obtains the vehicle model identification information.

[0040] In some embodiments, the in-vehicle terminal can send standardized diagnostic request commands to the standard interface of an on-board diagnostic (OBD) system via a communication connection to read vehicle identification number (VIN), model code, engine model, and other vehicle identification information. Furthermore, the in-vehicle terminal can store the read vehicle identification information in a local database after format verification for subsequent automatic vehicle identification to determine the vehicle's model category.

[0041] In some embodiments, the in-vehicle terminal can provide a three-level drop-down menu for brand, model series, and year through its user interface, while also including a vehicle identification number (VIN) input box. Users can then select or manually input relevant vehicle parameters, and the in-vehicle terminal can perform format validation on the user's input and extract complete vehicle identification information by matching it against its built-in vehicle information database.

[0042] In some embodiments, the vehicle terminal can capture the vehicle's license plate image and upload it to the cloud service platform. The cloud service platform can then extract the license plate number using optical character recognition technology, and call the vehicle registration database to match the corresponding vehicle identification number, vehicle model, manufacturing date, and other vehicle identification information. Finally, the complete vehicle identification information is sent back to the vehicle terminal as the recognition result.

[0043] In some embodiments, the vehicle terminal can capture images of the vehicle's model nameplate and use optical character recognition technology to extract identification fields such as vehicle identification number, model number, and curb weight marked on the nameplate, and integrate them into complete model identification information after rule verification.

[0044] In some embodiments, vehicle identification information may include the Vehicle Identification Number (VIN), also known as the VIN code, which is a 17-digit unique identifier for a vehicle and contains information such as manufacturer, model, and year. See the table below:

[0045] The first three digits of the VIN code are the World Manufacturer Identifier (WMI), used to identify the vehicle manufacturer. The fourth to eighth digits of the VIN code are the Vehicle Descriptor Section (VDS), used to describe features such as the model and engine. The tenth to seventeenth digits of the VIN code are the Vehicle Indicator Section (VIS), which includes the model year, factory code, and production serial number.

[0046] In some embodiments, the in-vehicle terminal may pre-store the vehicle identification code (VIN) locally, so that when the vehicle first starts the in-vehicle TTS application or when a new user first starts the in-vehicle TTS application, the identification code VIN can be directly read from the local storage to recommend in-vehicle audio to the user.

[0047] In some embodiments, the in-vehicle terminal can read the vehicle identification code (VIN) from the vehicle system through a pre-installed VIN reading module.

[0048] Step S102: Parse the vehicle identification information to obtain the vehicle type category.

[0049] It should be noted that vehicle type categories can include SPORT, BUSINESS, FAMILY, LUXURY, and EV. It should be understood that, based on different design needs in practical applications, vehicle type categories may also include other categories not listed here in different feasible implementations. The intelligent recommendation method for vehicle audio timbre provided in this application embodiment is not limited to a specific type of vehicle type.

[0050] After obtaining the vehicle's vehicle identification number (VIN), the vehicle terminal parses the VIN to extract the WMI (1-3 digits) to identify the manufacturer, the VDS (4-8 digits) to identify the vehicle's features, and the model year (10 digits) to identify the production year. Based on the manufacturer, vehicle features, and production year, the terminal queries the vehicle database to obtain the vehicle's vehicle category (SPORT / BUSINESS / FAMILY / LUXURY / EV).

[0051] In some embodiments, the vehicle terminal can parse the 17-digit VIN code of the vehicle based on a pre-set VIN parsing service, extract information such as manufacturer, vehicle series, model, and year from it, and query the model database to determine which of the vehicle's models is SPORT, BUSINESS, FAMILY, LUXURY, or EV.

[0052] Step S103: Construct a vehicle model profile based on the vehicle model category.

[0053] After obtaining the vehicle's model category, the in-vehicle terminal further constructs a model profile of the vehicle based on that model category.

[0054] In some embodiments, when constructing a vehicle model profile based on a vehicle type, the in-vehicle terminal can integrate relevant vehicle information (such as brand image, target user group, price range, and power type) corresponding to that vehicle type to construct the vehicle model profile.

[0055] In some embodiments, the vehicle terminal can also call a pre-trained vehicle model profile model (trained based on different vehicle categories and corresponding vehicle model profile labels), and input the vehicle's vehicle category into the model to obtain the vehicle model profile output by the model.

[0056] Step S104: Based on the cosine similarity between the vehicle model profile and multiple preset vehicle infotainment sounds, at least one first target vehicle infotainment sound is determined from the multiple preset vehicle infotainment sounds for recommendation processing.

[0057] It should be noted that the preset car infotainment system voice can be a variety of TTS voice resources stored in the vehicle's voice library, such as Xiao Ai (lively female voice), Xiao Shuai (sunny male voice), Xiao Chen (steady male voice), Yating (intellectual female voice), Xiao Mei (gentle female voice), Xiao Ming (friendly male voice), Ya Lin (elegant female voice), Jun Yi (magnetic male voice), and Xiao Zhi (neutral and technological voice), etc.

[0058] In some embodiments, each vehicle audio tone can be labeled with information such as the vehicle model category and scene it is compatible with.

[0059] After constructing the vehicle model profile, the in-vehicle terminal can further calculate the cosine similarity between the model profile and multiple preset vehicle infotainment sounds. Then, the in-vehicle terminal sorts the multiple preset vehicle infotainment sounds in descending order of cosine similarity and selects the top 1 to 3 first target vehicle infotainment sounds to recommend to the vehicle user (such as actively outputting the options of the 1 to 3 first target vehicle infotainment sounds in the TTS sound configuration interface for the user to confirm, etc.).

[0060] In some embodiments, when calculating the cosine similarity between the vehicle model profile and multiple preset vehicle audio timbres, the vehicle terminal may use a timbre-profile matching algorithm to calculate the cosine similarity between the timbre feature vector of each vehicle audio timbre and the feature vector of the vehicle model profile according to the following matching formula.

[0061] Similarity(V_voice, V_car) = (V_voice · V_car) / (‖V_voice‖ × ‖V_car‖).

[0062] Among them, V_voice is the timbre feature vector (including dimensions such as style, gender, and speech rate), and V_car is the vehicle model profile feature vector (including dimensions such as brand tone and target user group).

[0063] In some embodiments, the vector dimension of cosine similarity (Similarity) can be multiple (e.g., five dimensions: vitality, professionalism, warmth, technological feel, and elegance), with a value range of [0,1]. Based on this, the in-vehicle terminal can trigger tone recommendation when the cosine similarity (Similarity) is greater than the recommendation threshold (e.g., 0.7). For example, it can select the top 3 target in-vehicle audio tones with the highest cosine similarity from multiple preset in-vehicle audio to recommend to the vehicle's users.

[0064] In some embodiments, the vehicle terminal can also store the pre-calculated cosine similarity between different vehicle model profiles and multiple preset vehicle infotainment sounds locally. Then, after constructing the current vehicle model profile, it directly extracts the cosine similarity between the vehicle model profile and multiple preset vehicle infotainment sounds from the local storage, thereby selecting the first target vehicle infotainment sound with the top 1-3 cosine similarity to recommend to the vehicle's users.

[0065] In some embodiments, the in-vehicle terminal can also obtain the cosine similarity between the current vehicle model profile and multiple preset in-vehicle audio timbres from the vehicle system, and then select the first target in-vehicle audio timbres with the top 1-3 cosine similarity to recommend to the vehicle's users.

[0066] In this embodiment, when the vehicle-mounted terminal is in a cold start scenario for a new user / new vehicle using the in-vehicle TTS, in order to achieve personalized tone recommendations with zero manual operation by the user, it first obtains the vehicle model identification information, then parses the model identification information to obtain the vehicle model category (SPORT / BUSINESS / FAMILY / LUXURY / EV), and further constructs a vehicle model profile based on the model category. Finally, it calculates the cosine similarity between the model model profile and multiple preset vehicle-mounted tones. Then, the in-vehicle terminal sorts the multiple preset vehicle-mounted tones according to the cosine similarity from high to low, and recommends the top 1 to 3 first target vehicle-mounted tones to the vehicle user.

[0067] Thus, compared to the traditional solution that relies on users to perform tedious manual operations to configure the vehicle's TTS sound, the embodiments of this application can achieve personalized TTS sound recommendations for vehicle users with zero configuration. Without the need for users to manually configure the vehicle's sound, they can obtain a sound that matches the vehicle's tone, thereby solving the problem of the traditional vehicle TTS sound selection relying on manual selection by users and the tedious and complicated operation process, and improving the accuracy of sound recommendations.

[0068] Please refer to Figure 2 , Figure 2 for Figure 1 A detailed flowchart of step S103.

[0069] like Figure 2 As shown, in some embodiments, step S103 above: constructing a vehicle model profile based on the vehicle model category may include steps S201 and S202 as shown below.

[0070] Step S201: Determine the brand tone information of the vehicle from a preset brand tone database based on the vehicle type category.

[0071] It should be noted that the preset brand identity database can pre-set the identity tags of each brand and model, such as BMW = sporty + luxury, Volvo = safety + family, etc.

[0072] When building a vehicle profile based on vehicle type, the in-vehicle terminal can first query a pre-set brand tone database based on the vehicle type to determine the brand tone information of the vehicle.

[0073] Step S202: Based on the brand tone information and the target user group information, price range information and power type information corresponding to the vehicle type, construct the vehicle model profile.

[0074] After determining the brand and personality information of a vehicle, the in-vehicle terminal further combines this information with the target user group information (young / middle-aged / business professionals / families), price range information (economy / mid-range / high-end / luxury), and powertrain information (gasoline / hybrid / pure electric) corresponding to the vehicle's model category to construct a vehicle profile. For example, a vehicle's profile could be constructed by combining safety + family (brand personality), family users (user group), high-end (price), and hybrid (powertrain) to create "a hybrid high-end safe family car targeting family users."

[0075] Please refer to Figure 3 , Figure 3 for Figure 1 A detailed flowchart of step S104.

[0076] like Figure 3 As shown, in some embodiments, step S104 above: determining at least one first target vehicle audio tone from the multiple preset vehicle audio tones based on the cosine similarity between the vehicle model profile and multiple preset vehicle audio tones for recommendation processing may include steps S301 and S302 as shown below.

[0077] Step S301: Determine the target cosine similarity between the vehicle model profile and multiple preset vehicle system timbres from a preset vehicle model timbre mapping table; the vehicle model timbre mapping table includes: multiple vehicle model profiles, multiple preset vehicle system timbres, and the cosine similarity between the multiple vehicle model profiles and the multiple preset vehicle system timbres.

[0078] Step S302: Based on multiple target cosine similarities, at least one first target vehicle audio tone is determined for recommendation processing; among the multiple target cosine similarities, the cosine similarity between the first target vehicle audio tone and the vehicle model profile is higher than the cosine similarity between other vehicle audio tones besides the first target vehicle audio tone and the vehicle model profile.

[0079] The in-vehicle terminal can pre-calculate the cosine similarity between different vehicle model profiles and multiple preset vehicle infotainment system (VMS) sounds. It then sorts the cosine similarities and their corresponding preset VMS sounds in descending order, thus obtaining a vehicle model sound mapping table. Based on this, after parsing the vehicle's model identification information to obtain the vehicle's model category and constructing a vehicle model profile based on that category, the in-vehicle terminal can directly query the preset VMS sound mapping table to determine at least one first target VMS sound that has a mapping relationship with the model model profile. This allows for rapid recommendation of the first target VMS sound; specifically, it selects the top 1-3 cosine similarities from the current model model profile's cosine similarities as the recommended sound for the corresponding model category.

[0080] In some embodiments, the vehicle terminal can also pre-calculate the cosine similarity between different vehicle model profiles and multiple preset vehicle infotainment sounds, and then, based on the cosine similarity, select the first target vehicle infotainment sound with the top 1-3 cosine similarity as the recommended sound for the vehicle model category corresponding to each vehicle model profile, thereby establishing a mapping relationship between the first target vehicle infotainment sound and the vehicle model category, and obtaining a vehicle model sound mapping table between different vehicle models and different preset vehicle infotainment sounds.

[0081] In this way, the vehicle terminal can directly query the preset vehicle model sound mapping table based on the vehicle model category to determine at least one first target vehicle sound that has a mapping relationship with the vehicle model category, thereby enabling faster recommendation processing of the first target vehicle sound.

[0082] In some embodiments, the vehicle timbre mapping table pre-calculated by the in-vehicle terminal based on cosine similarity can be as follows:

[0083] Thus, after the vehicle terminal parses the vehicle identification information to determine that the vehicle type is family, it can use the vehicle type voice mapping table to determine the two voices, Xiaomei (gentle female voice) and Xiaoming (friendly male voice), as the first target vehicle voices to recommend to the user of the vehicle.

[0084] For example, the vehicle terminal can be configured as follows: Figure 4 The VIN parsing and vehicle model matching process shown involves reading the VIN code when the vehicle's infotainment system starts up. Then, the WMI (1-3 digits) in the VIN code is extracted to identify the manufacturer, the VDS (4-8 digits) in the VIN code is extracted to identify the vehicle model features, and the model year (10 digits) in the VIN code is extracted to identify the production year. The vehicle model database is then queried to obtain the vehicle model category (SPORT / BUSINESS / FAMILY / LUXURY / EV). Finally, based on the vehicle model category, a vehicle model timbre mapping table constructed by pre-calculating cosine similarity is queried to match the recommended timbre (i.e., the first target vehicle infotainment system timbre).

[0085] In this embodiment, a vehicle model profile is constructed based on the vehicle model category using an in-vehicle terminal. First, the brand tone information of the vehicle is determined based on the vehicle model category. Then, this brand tone information is combined with the vehicle's target user group information, price range information, and power type information to construct the vehicle model profile. Furthermore, the cosine similarity between the vehicle model profile and different in-vehicle infotainment system (IVS) timbres is calculated. For each vehicle model profile corresponding to its respective vehicle model category, the top 1-3 target IVS timbres with the highest cosine similarity are selected as the recommended timbres for that vehicle model category. This establishes a mapping relationship between the first target IVS timbres and the vehicle model category, resulting in a vehicle model timbre mapping table between different vehicle models and different preset IVS timbres. Therefore, the in-vehicle terminal can directly query the vehicle model timbre mapping table based on the vehicle model category to determine at least one first target IVS timbre with a mapping relationship to that vehicle model category for recommendation.

[0086] Thus, this embodiment can construct a multi-dimensional vehicle profile by deeply analyzing vehicle identification information, and combine it with a brand tone database to achieve personalized tone recommendations for new users with zero configuration. New users can obtain a tone that matches the vehicle tone when they use the vehicle's TTS-related applications for the first time, which effectively improves the accuracy of tone recommendations in the vehicle's TTS cold start scenario.

[0087] Please refer to Figure 5 , Figure 5 The following are flowcharts illustrating the steps involved in the vehicle audio intelligent recommendation method provided in some embodiments of this application.

[0088] like Figure 5 As shown, in some embodiments, the vehicle infotainment system audio recommendation method provided in this application, after step S104 above: determining at least one first target vehicle infotainment system audio based on the cosine similarity between the vehicle model profile and multiple preset vehicle infotainment system audios for recommendation processing, may further include steps S501 to S503 as shown below.

[0089] Step S501: Obtain the implicit behavioral information of the user of the vehicle regarding the first target vehicle infotainment system tone.

[0090] It should be noted that implicit behavioral information may include broadcast completion rate, volume adjustment, active switching, interruption frequency, etc.

[0091] In a cold start scenario where a new user / new car is using the in-vehicle TTS system, the in-vehicle terminal can recommend a first target in-vehicle sound tone that matches the vehicle's brand tone to the user. Furthermore, as the user continues to use the vehicle, the terminal can continuously acquire implicit behavioral information about the user's response to the first target in-vehicle sound tone.

[0092] In some embodiments, the vehicle terminal can continuously collect various implicit feedback signals as shown in the table below during the process of a user using the first target vehicle audio tone, thereby obtaining the implicit behavioral information of the user regarding the first target vehicle audio tone.

[0093]

[0094] Step S502: Perform multimodal fusion processing on the user's explicit preference information for vehicle audio and driving habit information, along with the implicit behavioral information and the time period features corresponding to the current time, to obtain a dynamic user profile.

[0095] It should be noted that the vehicle's audio preferences can include user-defined preferences for tone, speech rate, and volume. Driving habit information can include average speed (aggressive / smooth), common routes (commuting / long-distance), and driving time. The time period characteristics corresponding to the current time can include morning rush hour (concise and efficient), nighttime (gentle and soothing), and weekend (leisurely and lively).

[0096] After obtaining the implicit behavioral information of the user's preference for the first target vehicle audio tone, the vehicle terminal further performs multimodal fusion processing on the user's explicit preference information for the vehicle audio tone, the user's driving habits, the implicit behavioral information, and the time period characteristics corresponding to the current time. The fusion result of these multimodal information / data is then used as a dynamic user profile.

[0097] In some embodiments, when the vehicle terminal performs multimodal fusion on the user's self-set explicit preference information for vehicle audio timbre, the user's driving habit information for driving the vehicle, the user's implicit behavioral information on the first target vehicle audio timbre, and the time period characteristics corresponding to the current time, it can directly encapsulate these multimodal data into a data group, thereby using the data group as a dynamic user profile.

[0098] In some embodiments, the in-vehicle terminal can update the dynamic user profile based on the changed data when any of the continuously collected modal data changes during the user's use of the first target in-vehicle audio tone.

[0099] In some embodiments, when the in-vehicle terminal collects explicit audio preference information, driving habit information, implicit behavioral information, and time-period characteristics for multimodal fusion processing, it employs a confidence attenuation mechanism to exponentially decrease the confidence of historical data within these multimodal data. For example, the following attenuation formula is used to attenuate the confidence of historical data.

[0100] weight(t) = e^(-λt).

[0101] Where weight(t) is the confidence level of historical data for a certain modality, λ=0.1 / day, and t is the number of days from the current time.

[0102] Step S503: Based on the dynamic user profile, perform user preference update processing on the first target vehicle infotainment system sound to obtain the updated vehicle infotainment system sound.

[0103] After the vehicle terminal has built a dynamic user profile, it can update the user preference of the first target vehicle sound system based on the dynamic user profile to obtain the updated vehicle sound system, and then recommend the updated vehicle sound system to the user.

[0104] In some embodiments, when the vehicle terminal performs user preference update processing on the first target vehicle audio timbre based on the dynamic user profile, it can adjust the timbre parameters (such as fundamental frequency, speech rate, and volume) of the first target vehicle audio timbre to the timbre parameters corresponding to the dynamic user profile, thereby updating the first target vehicle audio timbre according to user preferences and obtaining the updated vehicle audio timbre.

[0105] In some embodiments, the in-vehicle terminal may directly use the updated vehicle audio tone, or it may use the updated vehicle audio tone only after obtaining further confirmation from the user regarding their recommendation.

[0106] In this embodiment, after recommending a first target in-vehicle infotainment sound tone that matches the vehicle's brand and style to the user via the in-vehicle terminal, the system continuously acquires the user's implicit behavioral information regarding the first target sound tone while the user continues to use the vehicle. It then performs multimodal fusion processing on the user's explicit sound tone preference information, driving habit information, implicit behavioral information, and time-segment characteristics corresponding to the current time to construct a dynamic user profile. Based on this dynamic user profile, the first target sound tone is updated to reflect user preferences, resulting in an updated sound tone. This approach not only enables personalized sound tone recommendations with zero user configuration but also further improves the accuracy of recommending in-vehicle infotainment sound to users by constructing a complete user profile based on multiple user preference sources (explicit settings, implicit behaviors, driving habits, usage time, etc.).

[0107] In some embodiments, the step S502 above, which involves "performing multimodal fusion processing of the user's explicit vehicle audio preference information, driving habit information, implicit behavioral information, and time period features corresponding to the current time," may include the following steps: Based on preset dynamic weights, the preference vectors of the vehicle infotainment system's explicit sound preference information, driving habit information, implicit behavioral information, and time period features are weighted and summed.

[0108] It should be noted that the dynamic weights include explicit preference weights, driving habit weights, implicit behavior weights, and time period feature weights. The explicit preference weights are greater than the driving habit weights, implicit behavior weights, and time period feature weights. The driving habit weights, implicit behavior weights, and time period feature weights are dynamically adjusted based on the data volume of the driving habit information, the implicit behavior information, and the time period feature weights, respectively.

[0109] The in-vehicle terminal can dynamically adjust the weights of the preference vectors for explicit audio preferences, driving habits, implicit behavioral information, and time-period features in advance, resulting in adjusted dynamic weights. For example, the explicit preference weight is at most 0.4, while other weights (driving habit weight, implicit behavioral weight, and time-period feature weight) are dynamically adjusted based on the data volume of each of the corresponding driving habit information, implicit behavioral information, and time-period feature. The ratio between these weights is directly proportional to the ratio between the data volume of each of these three elements.

[0110] Subsequently, when the in-vehicle terminal updates the user preference of the first target vehicle audio timbre based on the dynamic user profile, it can first perform a weighted summation of the preference vectors for the explicit preference information of the vehicle audio timbre, driving habit information, implicit behavioral information, and time-period characteristics, based on dynamic weights. In this way, the in-vehicle terminal can use the weighted summation result as the dynamic user profile or the user preference feedback corresponding to the dynamic user profile, and use it to update the user preference of the first target vehicle audio timbre.

[0111] In some embodiments, the vehicle terminal may use the following multimodal fusion formula to perform weighted summation of the preference vectors for the vehicle's audio timbre explicit preference information, driving habit information, implicit behavioral information, and time period features.

[0112] P_final = w1×P_explicit + w2×P_implicit + w3×P_driving + w4×P_time.

[0113] Where: w1=0.4 (explicit preference weight), w2=0.25 (implicit behavior weight), w3=0.2 (driving habit weight), w4=0.15 (time period feature weight), P_explicit, P_implicit, P_driving, and P_time are the preference vectors for each modality, with values ​​ranging from [0,1].

[0114] In some embodiments, when the in-vehicle terminal dynamically adjusts the weights of driving habits, implicit behavior, and time period features based on the data volume of each, if the data volume of any one of these modalities is insufficient (e.g., <10 samples), the weight of the missing modality can be proportionally allocated to the other modalities. For example, if the number of samples for driving habits is <10, the driving habits weight can be allocated to the implicit behavior weight and the time period feature weight according to the ratio between the data volume of the implicit behavior information and the time period feature.

[0115] For example, the vehicle-mounted terminal can adopt such as Figure 6 The user preference algorithm flow shown performs user preference update processing on the first target vehicle infotainment system audio timbre. Specifically, the in-vehicle terminal first performs implicit feedback collection to obtain implicit behavioral information of the vehicle user regarding the first target vehicle infotainment system audio timbre. Then, it combines the user's explicit audio timbre preference information, driving habit information, implicit behavioral information, and time-of-day characteristics to perform feature extraction processing, obtaining the preference features for each modality data in dimensions such as gender, style, and speech rate (where gender preference: male / female voice, value range [0,1], 0 indicates preference for male voice, 1 indicates preference for female voice; style preference: lively / calm, value range [0,1], 0 indicates preference for calm, 1 indicates preference for lively; speech rate preference: slow / fast, value range [0.5, 2.0]). 1.0 represents the preference vector at normal speaking speed. Then, the vehicle terminal uses the above-mentioned multimodal fusion formula to perform weighted summation on the preference vectors of the vehicle's timbre explicit preference information, driving habit information, implicit behavioral information, and time period features. The summation result (dynamic user profile / user preference feedback corresponding to the dynamic user profile) is substituted into the preference update formula shown below to calculate the preference. Finally, the vehicle terminal adjusts the relevant timbre parameters of the first target vehicle timbre based on the preference calculation result (updated preference value P_new), thereby completing the preference update of the first target vehicle timbre and obtaining the updated vehicle timbre.

[0116] P_new = α × P_current + (1 - α) × P_feedback.

[0117] Wherein, P_new is the updated preference value, ranging from [0,1], P_current is the current preference value, ranging from [0,1], and P_feedback is the current feedback value (i.e., the result of weighted summation of the preference vectors of the vehicle infotainment system's explicit sound preference information, driving habit information, implicit behavioral information, and time period features, P_final). Positive feedback is 1, negative feedback is 0, and the attenuation coefficient α = 0.8, indicating that historical preferences retain 80% weight, and new feedback accounts for 20% weight, making recent behavior have a greater impact on preferences.

[0118] In some embodiments, the in-vehicle terminal can also use a collaborative filtering recommendation mechanism to update the user preference of the first target in-vehicle audio tone based on the preferences of similar user groups (same vehicle model, similar usage habits) when the current user preference data (i.e., one or more of the explicit preference information of the in-vehicle audio tone, driving habit information, implicit behavioral information and time period characteristics) is insufficient, and then recommend the updated in-vehicle audio tone to the current user of the vehicle.

[0119] Please refer to Figure 7 , Figure 7 The following is a flowchart illustrating the steps of the vehicle audio intelligent recommendation method provided in some embodiments of this application.

[0120] like Figure 7 As shown, in some embodiments, the dynamic user profile includes the dynamic user profiles of multiple users of the vehicle. The vehicle audio intelligent recommendation method provided in this application embodiment may also include steps S701 to S703 as shown below.

[0121] Step S701: Obtain the user identification result of the vehicle.

[0122] When multiple users are present in a vehicle, the in-vehicle terminal can pre-recommend in-vehicle audio settings for each user and build a dynamic user profile for each user. Based on this dynamic user profile, it can then recommend in-vehicle audio settings that match the user's preferences. In this scenario, the in-vehicle terminal can identify the user using the vehicle and obtain the user identification result.

[0123] In some embodiments, the dynamic user profiles of multiple users are isolated from each other, that is, the vehicle terminal builds an independent preference profile for each user, and the user profiles do not interfere with each other.

[0124] In some embodiments, the vehicle terminal may identify the user of the vehicle through at least one of the following user identification methods.

[0125] Bluetooth phone pairing method: Detects the connected phone to identify the user using the vehicle; Seat memory position; inferring the vehicle user based on seat position; Voiceprint recognition (optional): Identifies the user of the vehicle via voice commands.

[0126] In some embodiments, when identifying vehicle users based on voiceprint recognition, the in-vehicle terminal can also directly construct a user profile based on speech recognition, and can make intelligent recommendations for in-vehicle audio timbres based on this user profile. For example, with the user's authorization to record, the in-vehicle terminal can analyze the characteristics of the user's voice (age, gender), and then combine these characteristics to recommend matching audio timbres to the user.

[0127] Step S702: Determine the target user profile from the multiple dynamic user profiles based on the user identification results.

[0128] After obtaining the user identification result of the vehicle, the vehicle terminal further determines the target user profile of the user indicated by the user identification result from a pre-built dynamic user profile of multiple users. For example, if the user identification result indicates user A, the vehicle terminal determines the dynamic user profile of A as the target user profile from the pre-built dynamic user profiles of users A, B, and C.

[0129] Step S703: Based on the target user profile, switch the second target vehicle infotainment sound to the third target vehicle infotainment sound; the second target vehicle infotainment sound includes the first target vehicle infotainment sound or the updated vehicle infotainment sound, and the third target vehicle infotainment sound is the vehicle infotainment sound after the first target vehicle infotainment sound has been updated with user preferences based on the target user profile.

[0130] After determining the target user profile, the in-vehicle terminal can switch the current second target in-vehicle audio tone (such as the first target in-vehicle audio tone recommended for the previous user, or the updated in-vehicle audio tone obtained after user preference update processing based on the dynamic user profile of the previous user) to the in-vehicle audio tone after user preference update processing of the first target in-vehicle audio tone using the target user profile.

[0131] In some embodiments, when the vehicle terminal detects a user switch in the vehicle and thus needs to use the target user profile of the switched user to switch the second target vehicle audio tone to the third target vehicle audio tone, the audio tone switch can be completed quickly within 500ms.

[0132] In some embodiments, the in-vehicle terminal can directly use the vehicle model's default tone when the vehicle is in guest mode (no user currently using the vehicle is identified).

[0133] In this embodiment, the vehicle's user is identified via the in-vehicle terminal to switch the vehicle's audio tone. This supports scenarios where multiple users use the vehicle (such as family-shared vehicles), automatically identifying the current user and switching the corresponding audio tone preference, thus solving the problem of traditional solutions not supporting personalized audio tone switching for multiple users. Furthermore, this embodiment can perform user identification through multiple methods or combinations of methods, ensuring audio tone switching while maintaining high accuracy in multi-user identification, with a switching latency of <500ms.

[0134] Please refer to Figure 8 , Figure 8 The following is a flowchart illustrating the steps of the vehicle audio intelligent recommendation method provided in some embodiments of this application.

[0135] like Figure 8As shown, in some embodiments, the vehicle audio intelligent recommendation method provided in this application may further include steps S801 and S802 as shown below.

[0136] Step S801: Obtain the usage scenario recognition result of the vehicle.

[0137] The in-vehicle terminal can also intelligently recommend audio settings based on the vehicle's usage scenarios. In this case, the in-vehicle terminal can continuously identify the vehicle's usage scenarios as the user operates the vehicle, thereby obtaining real-time usage scenario identification results for the vehicle.

[0138] In some embodiments, the vehicle's usage scenarios may include navigation mode, entertainment mode, vehicle control mode, call mode, and default mode, etc.

[0139] In some embodiments, when identifying the usage scenario of a vehicle, the in-vehicle terminal can identify the usage scenario of the vehicle based on the current App, vehicle status, and user behavior.

[0140] In some embodiments, the vehicle terminal may use the scene recognition rules shown in the table below to identify the vehicle’s usage scenario, and may simultaneously determine the recommended timbre characteristics under the usage scenario indicated by the usage scenario recognition result.

[0141]

[0142] Step S802: Based on the usage scenario recognition result, perform parameter gradual adjustment processing on the second target vehicle audio tone to switch the second target vehicle audio tone to the fourth target vehicle audio tone; the second target vehicle audio tone includes the first target vehicle audio tone or the updated vehicle audio tone, and the fourth target vehicle audio tone is the vehicle audio tone corresponding to the usage scenario recognition result.

[0143] It should be noted that the fourth target vehicle infotainment system voice tone corresponds to the usage scenario recognition result as follows: the fourth target vehicle infotainment system voice tone possesses the recommended voice tone characteristics for the usage scenario indicated by the usage scenario recognition result. For example, when the usage scenario recognition result indicates that the vehicle's usage scenario is navigation mode, the fourth vehicle infotainment system voice tone corresponding to the usage scenario recognition result can be a clear, concise, and moderately paced vehicle infotainment system voice tone.

[0144] After obtaining the vehicle usage scenario recognition result, the vehicle terminal can further adjust the parameters of the second target vehicle audio tone based on the vehicle usage scenario indicated by the usage scenario recognition result, so as to smoothly switch the second target vehicle audio tone to the fourth target vehicle audio tone.

[0145] In some embodiments, when the vehicle terminal switches the second target vehicle audio tone to the fourth target vehicle audio tone, it can use the following tone switching strategy to perform a smooth tone switching.

[0146]

[0147] Among them, the technical parameters related to crossfade can include: switching time 500ms, old tone 100%→0% (linear attenuation), new tone 0%→100% (linear enhancement).

[0148] In some embodiments, the vehicle terminal can pre-build a mapping relationship between different vehicle usage scenarios and different vehicle sound timbres, and each scenario can be associated with a primary timbre and alternative timbres.

[0149] In some embodiments, when the vehicle terminal switches the second target vehicle audio tone to the fourth target vehicle audio tone using a crossfading switching method, it may not directly replace the tone, but gradually adjust the tone parameters, including: fundamental frequency gradation: gradually changing from the fundamental frequency of the second target vehicle audio tone to the fundamental frequency of the fourth target vehicle audio tone within 500ms; speech rate gradation: adjusting the speech rate at the sentence boundary; and volume gradation, with a smooth transition within 200ms.

[0150] In some embodiments, when the vehicle terminal switches the second target vehicle audio tone to the fourth target vehicle audio tone, the switching can be performed at the end of the sentence or at a silence point (e.g., by gradually adjusting the tone parameters at the end of the sentence) to avoid switching in the middle of the sentence.

[0151] For example, the vehicle-mounted terminal can adopt such as Figure 9 The scene recognition and tone switching process shown is based on the recognition of the current usage scenario of the vehicle to realize the automatic switching of the vehicle's tone. The scene recognition and tone switching process includes three steps: scene detection, tone decision, and smooth switching. Scene detection can be performed according to the scene recognition rules mentioned above, tone decision can adopt the tone switching strategy mentioned above, and smooth switching adopts the cross-fading smooth switching mechanism mentioned above: the tone is quickly switched within 500ms, and the tone switching method is: the volume of the old tone (second target vehicle tone) gradually decreases from 100% to 0%, and the volume of the new tone (fourth target vehicle tone) gradually increases from 0% to 100%. In this way, the abruptness caused by the sudden change in tone can be avoided, and the user experience can be improved.

[0152] In this embodiment, the vehicle's usage scenario is identified through the in-vehicle terminal. Based on the identification results, the vehicle's audio parameters are gradually adjusted to switch the audio to a tone that is compatible with the vehicle's usage scenario. This not only enables intelligent switching of the vehicle's audio tone when the vehicle's usage scenario changes, but also ensures that the user is unaware of the tone change, achieving a smooth transition of tone with scene changes and improving the naturalness of the tone transition.

[0153] Next, we will present an embodiment of the vehicle audio timbre intelligent recommendation method provided in this application within an audio timbre recommendation system.

[0154] Please refer to Figure 10 , Figure 10 The system architecture diagram of the vehicle audio intelligent recommendation method provided in the embodiments of this application is shown in some embodiments.

[0155] like Figure 10 As shown, the voice recommendation system includes a cloud-based voice service layer and a vehicle-mounted client layer. The cloud-based voice service layer includes: a voice library storing various TTS voice resources, each voice labeled with its corresponding vehicle type and scenario; a VIN parsing service parsing the 17-digit VIN code to extract manufacturer, vehicle series, model, and year information; a user profile service storing and managing user preference data and supporting multi-user switching; and a recommendation engine that calculates the optimal voice recommendation based on vehicle model, user preferences, and scenario information. The vehicle-mounted client layer includes: a VIN reading module reading the vehicle's VIN code from the vehicle system; a scenario recognition module identifying the current usage scenario (navigation / entertainment / vehicle control / call); a behavior collection module collecting implicit user feedback (playback completion rate, volume adjustment, switching behavior); and a voice switching module executing voice switching, supporting crossfade smooth transitions.

[0156] The data flow of the tone recommendation system includes: the vehicle client reads the VIN → the cloud VIN parsing service → returns vehicle model information and the cloud recommendation engine → returns the recommended tone list → the vehicle client tone switching module and vehicle client behavior collection → the cloud user profile service → updates user preferences.

[0157] In some embodiments, the data structure for vehicle model information exchanged between the cloud-based timbre service layer and the vehicle-mounted client layer in the timbre recommendation system can be: struct VehicleInfo { string vin; / / 17-digit VIN code string manufacturer; / / Manufacturer string model_series; / / Car series string model_name; / / Model name int model_year; / / Year model VehicleType type; / / Vehicle type: SPORT / BUSINESS / FAMILY / LUXURY / EV }

[0158] The data structure for the audio information exchanged between the cloud-based audio service layer and the in-vehicle client layer can be: struct VoiceProfile { string voice_id; / / Unique identifier for voice timbre string voice_name; / / Voice name Gender; / / Gender: MALE / FEMALE Style style; / / Style: LIVELY / STEADY / PROFESSIONAL / WARM VehicleType[] suitable_types; / / Suitable vehicle types Scenario[] suitable_scenarios; / / Suitable scenarios }

[0159] Furthermore, the data structure for user preferences interacting between the cloud-based audio service layer and the in-vehicle client layer can be: struct UserPreference { string user_id; / / User identifier Gender preferred_gender; / / Preferred gender Style preferred_style; / / Style preference float preferred_speed; / / Preferred speaking speed (0.5-2.0) map<Scenario, string> scenario_voice_map; / / Scene-voice mapping long last_update_time; / / Last update time }

[0160] In this embodiment, the timbre recommendation system achieves a comprehensive performance and experience upgrade compared to traditional solutions. In terms of user operation, it abandons the traditional cumbersome manual setup mode, achieving fully automated operation with zero configuration. User configuration is 100% simplified, significantly lowering the barrier to entry. Furthermore, regarding timbre recommendation accuracy, the cold start phase uses intelligent recommendation based on VIN vehicle model profiles, replacing the traditional fixed default timbre, effectively improving the matching accuracy between timbre and vehicle model. Moreover, with long-term user use, the matching accuracy can be further improved, thus increasing user satisfaction. Moreover, by rapidly adjusting timbre parameters to smoothly switch timbre according to scene changes, it solves the shortcomings of traditional solutions where timbre switching is abrupt or non-existent. Finally, the system also supports automatic multi-user identification and independent profiles, comprehensively capturing user preferences through multimodal fusion, significantly improving scene and user adaptability.

[0161] Please refer to Figure 11 This application also provides an intelligent recommendation system for vehicle infotainment system audio timbre, which can implement the above-mentioned intelligent recommendation method for vehicle infotainment system audio timbre.

[0162] like Figure 11 As shown in the embodiments of this application, the intelligent recommendation system for vehicle audio timbre may include: The acquisition module is used to acquire vehicle model identification information; The vehicle identification information parsing module is used to parse and process the vehicle identification information to obtain the vehicle type. The vehicle model profile building module is used to build a vehicle model profile based on the vehicle model category; The intelligent timbre recommendation module is used to determine at least one first target vehicle timbre from the multiple preset vehicle timbres based on the cosine similarity between the vehicle model profile and multiple preset vehicle timbres for recommendation processing.

[0163] In some embodiments, the vehicle model profile building module is further configured to determine the brand tone information of the vehicle from a preset brand tone database based on the vehicle model category; and to build a vehicle model profile based on the brand tone information and the target user group information, price range information and power type information corresponding to the vehicle model category.

[0164] In some embodiments, the intelligent timbre recommendation module is further configured to determine the target cosine similarity between the vehicle model profile and multiple preset vehicle infotainment timbres from a preset vehicle model timbre mapping table; the vehicle model timbre mapping table includes: multiple vehicle model profiles, multiple preset vehicle infotainment timbres, and cosine similarity between the multiple vehicle model profiles and the multiple preset vehicle infotainment timbres; at least one first target vehicle infotainment timbre is determined for recommendation processing based on the multiple target cosine similarities; among the multiple target cosine similarities, the cosine similarity between the first target vehicle infotainment timbre and the vehicle model profile is higher than the cosine similarity between other vehicle infotainment timbres (excluding the first target vehicle infotainment timbre) and the vehicle model profile.

[0165] In some embodiments, the in-vehicle audio tone intelligent recommendation system provided in this application may further include: The multimodal preference fusion module is used to obtain implicit behavioral information of the user of the vehicle regarding the first target vehicle infotainment system tone; to perform multimodal fusion processing on the user's explicit vehicle infotainment system tone preference information, driving habit information, implicit behavioral information, and time period features corresponding to the current time to obtain a dynamic user profile; and to perform user preference update processing on the first target vehicle infotainment system tone based on the dynamic user profile to obtain an updated vehicle infotainment system tone.

[0166] In some embodiments, the multimodal preference fusion module is further configured to perform weighted summation processing on the preference vectors of the vehicle infotainment system's explicit timbre preference information, the driving habit information, the implicit behavior information, and the time period feature, based on preset dynamic weights; the dynamic weights include explicit preference weights, driving habit weights, implicit behavior weights, and time period feature weights, wherein the explicit preference weights are greater than the driving habit weights, implicit behavior weights, and time period feature weights, and the driving habit weights, implicit behavior weights, and time period feature weights are dynamically adjusted based on the data volume of the driving habit information, the implicit behavior information, and the time period feature.

[0167] In some embodiments, the dynamic user profile includes dynamic user profiles of multiple users of the vehicle. The in-vehicle audio intelligent recommendation system provided in this application embodiment may further include: A multi-user identification and personalization switching module is used to obtain the user identification result of the vehicle; determine the target user profile from multiple dynamic user profiles based on the user identification result; and switch the second target vehicle audio tone to the third target vehicle audio tone based on the target user profile; the second target vehicle audio tone includes the first target vehicle audio tone or the updated vehicle audio tone, and the third target vehicle audio tone is the vehicle audio tone after user preference updating processing of the first target vehicle audio tone based on the target user profile.

[0168] In some embodiments, the in-vehicle audio tone intelligent recommendation system provided in this application may further include: A scene-aware tone smoothing switching module is used to obtain the vehicle's usage scene recognition result; based on the usage scene recognition result, the second target vehicle tone is subjected to parameter gradual adjustment processing to switch the second target vehicle tone to a fourth target vehicle tone; the second target vehicle tone includes the first target vehicle tone or the updated vehicle tone, and the fourth target vehicle tone is the vehicle tone corresponding to the usage scene recognition result.

[0169] It should be noted that the specific implementation of the vehicle infotainment system with intelligent audio recommendation provided in this application is basically the same as the specific implementation of the vehicle infotainment system with intelligent audio recommendation method described above, and will not be repeated here.

[0170] Please see Figure 12 This application also provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the above-mentioned intelligent recommendation method for vehicle audio timbre.

[0171] In some embodiments, the electronic device may be any smart terminal such as a vehicle, an in-vehicle terminal, an in-vehicle hardware platform (e.g., an in-vehicle computer), a tablet computer, a smartphone, or a wearable device; or, the electronic device may be a vehicle including a memory and a processor.

[0172] like Figure 12 As shown, the electronic device provided in this application embodiment may include: The processor 1201 can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this application. The memory 1202 can be implemented in the form of read-only memory (ROM), static storage device, dynamic storage device, or random access memory (RAM). The memory 1202 can store the operating system and other applications. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory 1202 and is called and executed by the processor 1201 to execute the vehicle audio intelligent recommendation method of the embodiments of this application. The input / output interface 1203 is used to implement information input and output; The communication interface 1204 is used to enable communication and interaction between this device and other devices. Communication can be achieved through wired means (such as USB, network cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.). Bus 1205 transmits information between various components of the device (e.g., processor 1201, memory 1202, input / output interface 1203, and communication interface 1204); The processor 1201, memory 1202, input / output interface 1203 and communication interface 1204 are connected to each other within the device via bus 1205.

[0173] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described intelligent recommendation method for vehicle audio timbre.

[0174] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, memory may optionally include memory remotely located relative to the processor, and these remote memories can be connected to the processor via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0175] This application also provides a computer program product, including a computer program. The steps implemented by the computer program when executed by a processor are basically the same as those in the specific embodiments of the above-described vehicle audio intelligent recommendation method, and will not be repeated here.

[0176] The embodiments described in this application are for the purpose of more clearly illustrating the technical solutions of this application, and do not constitute a limitation on the technical solutions provided in this application. As those skilled in the art will know, with the evolution of technology and the emergence of new application scenarios, the technical solutions provided in this application are also applicable to similar technical problems.

[0177] Those skilled in the art will understand that the technical solutions shown in the figures do not constitute a limitation on the embodiments of this application, and may include more or fewer steps than shown, or combine certain steps, or different steps.

[0178] The system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0179] Those skilled in the art will understand that all or some of the steps in the methods disclosed above, as well as the functional modules / units in the systems and devices, can be implemented as software, firmware, hardware, or suitable combinations thereof.

[0180] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0181] It should be understood that in this application, "at least one (item)" means one or more, and "more than" means two or more. "And / or" is used to describe the relationship between related objects, indicating that three relationships can exist. For example, "A and / or B" can represent: only A exists, only B exists, and both A and B exist simultaneously, where A and B can be singular or plural. The character " / " generally indicates that the preceding factor is divided by the following factor, or that the related objects are in an "or" relationship. "At least one (item) of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one (item) of a, b, or c can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0182] In the embodiments provided in this application, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of the units described above is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between systems or units may be electrical, mechanical, or other forms.

[0183] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0184] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0185] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes multiple instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing programs, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0186] The preferred embodiments of the present application have been described above with reference to the accompanying drawings, but this does not limit the scope of the claims of the present application. Any modifications, equivalent substitutions, and improvements made by those skilled in the art without departing from the scope and substance of the embodiments of the present application shall be within the scope of the claims of the present application.

Claims

1. A method for intelligent recommendation of in-vehicle audio timbre, characterized in that, The method includes: Obtain vehicle model identification information; The vehicle identification information is parsed to obtain the vehicle type category; Construct a vehicle model profile based on the vehicle model category; Based on the cosine similarity between the vehicle model profile and multiple preset vehicle infotainment system tones, at least one first target vehicle infotainment system tones is determined from the multiple preset vehicle infotainment system tones for recommendation processing.

2. The method according to claim 1, characterized in that, The process of constructing a vehicle model profile based on the vehicle model category includes: The brand tone information of the vehicle is determined from a preset brand tone database based on the vehicle type category; Based on the brand image information and the target user group information, price range information, and power type information corresponding to the vehicle type, a vehicle profile is constructed.

3. The method according to claim 1, characterized in that, The step of determining at least one first target vehicle infotainment sound from the multiple preset vehicle infotainment sound sources based on the cosine similarity between the vehicle model profile and multiple preset vehicle infotainment sound sources for recommendation processing includes: The target cosine similarity between the vehicle model profile and multiple preset vehicle infotainment system timbres is determined from a preset vehicle model timbre mapping table; the vehicle model timbre mapping table includes: multiple vehicle model profiles, multiple preset vehicle infotainment system timbres, and the cosine similarity between the multiple vehicle model profiles and the multiple preset vehicle infotainment system timbres. At least one first target vehicle infotainment sound tone is determined based on multiple target cosine similarities for recommendation processing; among the multiple target cosine similarities, the cosine similarity between the first target vehicle infotainment sound tone and the vehicle model profile is higher than the cosine similarity between other vehicle infotainment sound tones and the vehicle model profile.

4. The method according to claim 1, characterized in that, After performing the recommendation process by determining at least one first target vehicle audio tone from the plurality of preset vehicle audio tones, the method further includes: Obtain implicit behavioral information about the user's response to the first target vehicle infotainment system sound. The explicit audio preferences and driving habits of the vehicle's users are combined with the implicit behavioral information and the time period features corresponding to the current time to form a dynamic user profile through multimodal fusion processing. Based on the dynamic user profile, the first target vehicle infotainment system's audio timbre is updated according to user preferences to obtain the updated audio timbre.

5. The method according to claim 4, characterized in that, The process of multimodal fusion of the user's explicit audio preferences and driving habits with the implicit behavioral information and the time-segment features corresponding to the current time includes: Based on preset dynamic weights, the preference vectors of the vehicle infotainment system's explicit sound preference information, driving habit information, implicit behavioral information, and time period features are weighted and summed. The dynamic weights include explicit preference weights, driving habit weights, implicit behavior weights, and time period feature weights. The explicit preference weights are greater than the driving habit weights, implicit behavior weights, and time period feature weights. The driving habit weights, implicit behavior weights, and time period feature weights are dynamically adjusted based on the data volume of the driving habit information, the implicit behavior information, and the time period feature, respectively.

6. The method according to claim 4, characterized in that, The dynamic user profile includes dynamic user profiles for multiple users of the vehicle, and the method further includes: Obtain the user identification result of the vehicle; Based on the user identification results, a target user profile is determined from multiple dynamic user profiles; Based on the target user profile, the second target vehicle infotainment system audio is switched to the third target vehicle infotainment system audio; the second target vehicle infotainment system audio includes the first target vehicle infotainment system audio or the updated vehicle infotainment system audio, and the third target vehicle infotainment system audio is the vehicle infotainment system audio after the first target vehicle infotainment system audio is updated based on the target user profile by performing user preference updates.

7. The method according to claim 1 or 4, characterized in that, The method further includes: Obtain the usage scenario recognition results of the vehicle; Based on the usage scenario recognition result, the second target vehicle infotainment system audio timbre is subjected to parameter gradual adjustment processing to switch the second target vehicle infotainment system audio timbre to the fourth target vehicle infotainment system audio timbre; the second target vehicle infotainment system audio timbre includes the first target vehicle infotainment system audio timbre or the updated vehicle infotainment system audio timbre, and the fourth target vehicle infotainment system audio timbre is the vehicle infotainment system audio timbre corresponding to the usage scenario recognition result.

8. An electronic device, characterized in that, The electronic device includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements the vehicle audio intelligent recommendation method according to any one of claims 1 to 7.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the vehicle infotainment system intelligent sound recommendation method as described in any one of claims 1 to 7.

10. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, implements the vehicle infotainment system audio intelligent recommendation method as described in any one of claims 1 to 7.

Images