Processing method and device for voice broadcast of vehicle-mounted terminal, and vehicle

By judging the intent and functional domain of the voice command on the vehicle terminal and using preset tags to decide whether to call the cloud language generation model, the problem that the vehicle terminal cannot efficiently provide high-quality voice broadcasts, and achieve efficient and high-quality voice broadcasts. Improved user experience.

CN119091851APending Publication Date: 2024-12-06CHONGQING CHANGAN TECH CO LTD
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202411221068.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-09-02
Publication Date
2024-12-06

AI Technical Summary

Technical Problem

The vehicle-machine terminal cannot efficiently provide high-quality voice broadcasts, and the large-model-generated voice packets consume high resources in high concurrency scenarios, which has affected the user experience and has not been mass-produced and used in the on-board field.

Method used

By judging the intent and functional domain of the voice command on the vehicle terminal, using preset tags to determine whether to call the cloud language generation model, efficient voice broadcasting is achieved, and high-quality voice broadcasting is provided when needed, reducing resource consumption and latency impacts.

Benefits of technology

It improves the efficiency and quality of voice broadcasting on the vehicle and computer side, meets users' high-quality interaction needs, reduces the experience problems caused by resource consumption and delays, and improves user experience and satisfaction.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119091851A_ABST
    Figure CN119091851A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of vehicle technology voice broadcast, and discloses a processing method and device for voice broadcast of a vehicle infotainment terminal and an automobile, and the method comprises the steps: obtaining a voice instruction of a user, and determining a user instruction intention according to the voice instruction; judging a falling domain type of a functional domain to which the voice instruction belongs according to the user instruction intention, and determining a preset label of the falling domain type according to a functional domain label constraint rule; and judging whether to call a language generation model pre-constructed by the cloud based on the preset tag, so as to carry out voice broadcasting according to the user instruction intention. According to the invention, label constraint is carried out on the function domain, whether the language generation model needs to be called to carry out voice broadcast can be determined according to the user instruction intention, efficient voice broadcast is realized when calling is not needed, and high-quality voice broadcast is provided when calling is needed, so that efficient and high-quality voice broadcast is provided for a user at a vehicle terminal, and the user experience is improved. And the user experience and the satisfaction degree are improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle technology voice broadcasting, and in particular to a method and device for processing vehicle-side voice broadcasting, and a vehicle. Background Art

[0002] Voice packages generated by large model technology are increasingly widely used in life because they are more emotional, and their timbre and naturalness are closer to real people. Various AI voice packages and AI chat software have been well received by users, gradually forming a new track with huge potential market value. However, since large models consume a lot of resources to generate voice packages and broadcast them in real time, especially in high-concurrency usage scenarios, the high cost makes car companies dare not take risks; and large models need to be generated in the cloud, which will have a delay of 1s-2s, which will also have a certain impact on the user experience. Therefore, the voice packages generated by large models have not been officially mass-produced and used in the automotive field, making it impossible for the car machine to efficiently provide higher-quality voice broadcasts. Summary of the invention

[0003] In view of this, the present invention provides a method, device and vehicle for processing voice broadcast on a vehicle terminal, so as to solve the problem that the vehicle terminal cannot efficiently provide high-quality voice broadcast.

[0004] In a first aspect, the present invention provides a method for processing a voice broadcast on a vehicle terminal, the method comprising:

[0005] Obtain the user's voice command and determine the user's command intention based on the voice command;

[0006] Determine the domain type of the functional domain to which the voice command belongs according to the user's command intention, and determine the preset label of the domain type according to the functional domain label constraint rule;

[0007] Based on the preset tags, it is determined whether to call the language generation model pre-built in the cloud to perform voice broadcast according to the user's command intention.

[0008] The method for processing the voice broadcast on the vehicle side provided by the embodiment of the present invention determines the user's command intention according to the user's voice command, determines the domain type of the functional domain to which the voice command belongs according to the user's command intention, and determines the corresponding preset label according to the functional domain label constraint rule, and determines whether to call the language generation model pre-built in the cloud based on the preset label to perform voice broadcast according to the user's command intention. By constraining the functional domain with labels, the present invention can determine whether it is necessary to call the language generation model for voice broadcast according to the user's command intention, achieve efficient voice broadcast when no call is needed, and provide high-quality voice broadcast when call is needed, thereby providing users with efficient and high-quality voice broadcast on the vehicle side, improving user experience and satisfaction.

[0009] In an optional implementation, determining the user instruction intention based on the voice instruction includes: performing voice recognition on the voice instruction to obtain voice text information; and parsing the voice text information based on preset text matching rules to determine the user instruction intention.

[0010] By judging the intention of user voice, the present invention can accurately grasp user needs, thereby providing users with accurate services and improving user satisfaction.

[0011] In an optional implementation, the domain type of the functional domain to which the voice command belongs is determined according to the user's command intention, and a preset label of the domain type is determined according to the functional domain label constraint rules, including: screening the preset application corresponding to the voice command in the vehicle-mounted functional applications on the vehicle-mounted side based on the user's command intention; determining the domain type of the functional domain to which the voice command belongs according to the preset application; obtaining the preset functional domain label constraint rules, and determining the preset label of the domain type according to the functional domain label constraint rules, the preset labels including: fixed reply labels and real-time reply labels.

[0012] By performing label constraints on different functional domains, the present invention can differentiate the reply content based on the usage scenario, thereby achieving fixed replies for some high-frequency and repetitive scenarios and real-time replies for the remaining scenarios. At the same time, the high efficiency and high quality of voice broadcasting are met, and resource consumption and experience problems caused by high latency are reduced.

[0013] In an optional embodiment, it is determined based on a preset tag whether it is necessary to call a language generation model pre-built in the cloud to perform voice broadcast according to the user's command intention, including: if the preset tag is a fixed reply tag, then judging whether the local voice resources contain the corresponding first voice package based on the user's command intention and the local voice resources, if so, broadcasting the first voice package; if not, sending a resource call instruction to the cloud, so that the cloud determines whether the cloud voice resources contain the corresponding second voice package based on the user's command intention and the cloud voice resources, if so, broadcasting the second voice package sent by the cloud, and storing the second voice package in the local voice resources; if not, sending a model call instruction to the cloud, so that the cloud generates the corresponding third voice package based on the user's command intention and the language generation model, broadcasting the third voice package sent by the cloud, and storing the third voice package in the local voice resources.

[0014] The present invention caches and then broadcasts some high-frequency fixed replies based on a language generation model, and can directly call local voice resources for efficient voice broadcasting. At the same time, based on a cloud-based language generation model, the local voice resources are continuously enriched to provide high-quality voice broadcasting for home use, which can not only reduce resource consumption on the vehicle side, but also improve the experience problem caused by voice generation delay.

[0015] In an optional embodiment, based on the preset tag, it is determined whether it is necessary to call the language generation model pre-built in the cloud to perform voice broadcast according to the user's command intention, and also includes: if the preset tag is a real-time reply tag, a model call instruction is sent to the cloud to enable the cloud to generate a corresponding fourth voice package based on the user's command intention and the language generation model, and broadcast the fourth voice package sent by the cloud.

[0016] The present invention can meet the user's usage needs by performing voice generation and broadcasting in a real-time reply scenario, provide the user with high-quality voice broadcasting under the condition of limited resources on the vehicle side, and improve the user's voice interaction experience.

[0017] In the second aspect, the present invention provides a device for processing voice broadcast on a vehicle side, the device comprising: an intention determination module, used to obtain the user's voice command and determine the user's command intention based on the voice command; a label judgment module, used to judge the domain type of the functional domain to which the voice command belongs based on the user's command intention, and determine the preset label of the domain type based on the functional domain label constraint rules; a voice broadcast module, used to judge whether it is necessary to call a language generation model pre-built in the cloud based on the preset label to perform voice broadcast according to the user's command intention.

[0018] In a third aspect, the present invention provides a computer device, comprising: a memory and a processor, the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the method for processing vehicle-side voice broadcast according to the first aspect or any corresponding embodiment thereof by executing the computer instructions.

[0019] In a fourth aspect, the present invention provides a computer-readable storage medium having computer instructions stored thereon, the computer instructions being used to enable a computer to execute the method for processing vehicle-side voice broadcast according to the first aspect or any corresponding embodiment thereof.

[0020] In a fifth aspect, the present invention provides a computer program product, including computer instructions, which are used to enable a computer to execute the method for processing vehicle-side voice broadcast according to the first aspect or any corresponding embodiment thereof.

[0021] In a sixth aspect, the present invention provides a car, comprising: a storage unit for storing local voice resources; a controller connected to the storage unit, for executing the method for processing the vehicle-side voice broadcast of the above-mentioned first aspect or any corresponding embodiment thereof.

[0022] Beneficial effects of the present invention:

[0023] (1) By deploying a speech generation model in the cloud, the present invention can generate a response with richer emotions, a timbre and naturalness closer to that of a real person according to the user's voice command, thereby satisfying high-quality voice interaction with the user;

[0024] (2) The present invention can distinguish the content of the reply words based on the usage scenario by labeling the functional domain to which the voice command belongs. Some high-frequency, repetitive, and commonly used fixed reply words are cached before being broadcast, thereby improving the broadcast speed of the voice reply and reducing the resource consumption of the voice broadcast on the vehicle side. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0026] Figure 1 is a flow chart of a method for processing a voice announcement on a vehicle terminal according to an embodiment of the present invention;

[0027] Figure 2 is a flow chart of another vehicle-side voice broadcast processing system according to an embodiment of the present invention;

[0028] Figure 3 is a flow chart of another method for processing a vehicle-side voice broadcast according to an embodiment of the present invention;

[0029] Figure 4 is a schematic diagram of a judgment flow chart of another method for processing a vehicle-side voice broadcast according to an embodiment of the present invention;

[0030] Figure 5 is a flow chart of another method for processing voice broadcast at a vehicle terminal according to an embodiment of the present invention;

[0031] Figure 6 is a structural block diagram of a device for processing voice broadcast at a vehicle terminal according to an embodiment of the present invention;

[0032] Figure 7 It is a schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0033] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0034] The embodiment of the present invention is applicable to the scenario of voice interaction according to the driver's voice command during the driver's driving process. The embodiment of the present invention provides a method for processing voice broadcast on a vehicle side, by judging the domain type of the functional domain to which the voice command belongs, and determining whether to call a language generation model according to the label of the domain type to achieve the effect of efficient and high-quality voice broadcast.

[0035] According to an embodiment of the present invention, an embodiment of a method for processing voice broadcast on a vehicle terminal is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0036] In this embodiment, a method for processing a voice broadcast on a vehicle terminal is provided, which can be used in a controller of a vehicle. Figure 1 is a flow chart of a method for processing a vehicle-side voice broadcast according to an embodiment of the present invention. Figure 1 As shown, the process includes the following steps:

[0037] Step S101, obtaining a user's voice command, and determining the user's command intention according to the voice command.

[0038] Specifically, in an embodiment of the present invention, a voice collection device, such as a microphone, is deployed on the car to collect the voice commands of the user in the car while the driver is driving the vehicle. For example, the driver says "I want to go to Guanyin Bridge", "I want to listen to music", "What's the weather like recently" and so on. When the user issues a voice command, the voice command collected by the voice collection device is obtained, and the user's command intention is determined according to the voice command. For example, "I want to go to Guanyin Bridge" corresponds to the navigation intention, and "I want to listen to music" corresponds to the music playback intention, which is only an example and is not limited to this.

[0039] Step S102: determining the domain type of the functional domain to which the voice command belongs according to the user's command intention, and determining a preset label of the domain type according to the functional domain label constraint rule.

[0040] Specifically, in an embodiment of the present invention, the functional domains of the vehicle terminal are labeled and constrained in advance based on the user's usage scenarios, wherein the labels include fixed reply labels and real-time reply labels, and fixed functional domain label constraint rules are obtained. Among them, fixed reply labels are applied to the functional domains of vehicle functions corresponding to high-frequency, repetitive, and commonly used scenarios, such as navigation functional domains, music playback functional domains, etc., which are fixed reply labels; real-time reply labels are applied to question-and-answer and chat functions. In an embodiment of the present invention, after the driver issues a voice command while driving the vehicle, the domain type of the functional domain to which the voice command belongs is determined according to the user's command intent corresponding to the voice command, such as navigation functional domains, music playback functional domains, window control functional domains, etc., and then the preset label corresponding to the current domain type is determined to be a fixed reply label or a real-time reply label according to the functional domain label constraint rules.

[0041] Step S103, based on the preset tags, determines whether it is necessary to call the language generation model pre-built in the cloud to perform voice broadcast according to the user's instruction intention.

[0042] Specifically, in an embodiment of the present invention, a language generation model is pre-constructed in the cloud, wherein the language generation model is a machine learning model with large-scale parameters and complex computing structure, called a large model, which can process massive data and complete various complex tasks. In the field of natural language processing, it is a large language model that can be trained with a large amount of data and parameters to generate human-like text or answer natural language questions, and provide users with voice responses with fuller emotions, timbre and naturalness closer to real people. Due to its high resource consumption, it cannot be used in mass production on the vehicle side, so it is deployed in the cloud. The vehicle side can communicate with the cloud, so as to send the user's voice command to the cloud, obtain the voice reply generated by the cloud according to the voice command, and play it. However, due to the delay of 1s-2s in voice generation, the user experience is affected.

[0043] In some optional implementations, the embodiments of the present invention store some high-frequency and repeatedly used voice packages generated by the large language model as local voice resources for efficient voice broadcasting. For those that do not exist in the local voice resources, it is necessary to call the large language model in the cloud to generate voice. The specific generation process is a conventional technical means in this field and will not be repeated here. The embodiment of the present invention determines whether it is necessary to call the large language model for voice broadcasting based on the preset label corresponding to the domain type of the functional domain to which the voice instruction belongs. If it is a fixed reply label, the corresponding voice package in the local voice resource is broadcast. If it is a real-time reply label, the large language model is called to generate a real-time voice package and broadcast it.

[0044] The method for processing the voice broadcast on the vehicle side provided by the embodiment of the present invention determines the user's command intention according to the user's voice command, determines the domain type of the functional domain to which the voice command belongs according to the user's command intention, and determines the corresponding preset label according to the functional domain label constraint rule, and determines whether to call the language generation model pre-built in the cloud based on the preset label to perform voice broadcast according to the user's command intention. By constraining the functional domain with labels, the present invention can determine whether it is necessary to call the language generation model for voice broadcast according to the user's command intention, achieve efficient voice broadcast when no call is needed, and provide high-quality voice broadcast when call is needed, thereby providing users with efficient and high-quality voice broadcast on the vehicle side, improving user experience and satisfaction.

[0045] In this embodiment, a method for processing a voice broadcast on a vehicle terminal is provided, which can be used in a controller of a vehicle. Figure 2 is a flow chart of a method for processing a vehicle-side voice broadcast according to an embodiment of the present invention. Figure 2 As shown, the process includes the following steps:

[0046] Step S201, obtaining a user's voice command, and determining the user's command intention according to the voice command.

[0047] Specifically, the above step S201 includes:

[0048] Step S2011, perform voice recognition on the voice command to obtain voice text information.

[0049] Specifically, in the embodiment of the present invention, the vehicle terminal cannot process the voice command in the form of audio, so it is necessary to perform voice recognition on the voice command to obtain processable voice text information. Among them, voice recognition is a conventional technical means in this field and will not be described in detail here.

[0050] Step S2012, parsing the voice text information based on preset text matching rules to determine the user's instruction intention.

[0051] Specifically, in the embodiment of the present invention, regular expressions are used as text matching rules, wherein regular expressions are a powerful text processing tool for matching, searching, replacing and extracting information in character strings, and can find qualified text in the text by parsing the voice text information. For example, if the user's voice command is "I want to go to Guanyin Bridge", the qualified command sentence "I want to go" and the noun word slot "Guanyin Bridge" can be found through regular expressions, thereby determining that the user's command intention is a navigation intention.

[0052] Step S202: determine the domain type of the functional domain to which the voice command belongs according to the user's command intention, and determine the preset label of the domain type according to the functional domain label constraint rule. Figure 1Step S102 of the illustrated embodiment will not be described in detail here.

[0053] Step S203, based on the preset tags, determines whether it is necessary to call the language generation model pre-built in the cloud to perform voice broadcast according to the user's instruction intention. Figure 1 Step S103 of the illustrated embodiment will not be described in detail here.

[0054] The method for processing the voice broadcast on the vehicle side provided by the embodiment of the present invention determines the user's command intention according to the user's voice command, determines the domain type of the functional domain to which the voice command belongs according to the user's command intention, and determines the corresponding preset label according to the functional domain label constraint rule, and determines whether to call the language generation model pre-built in the cloud based on the preset label to perform voice broadcast according to the user's command intention. By constraining the functional domain with labels, the present invention can determine whether it is necessary to call the language generation model for voice broadcast according to the user's command intention, achieve efficient voice broadcast when no call is needed, and provide high-quality voice broadcast when call is needed, thereby providing users with efficient and high-quality voice broadcast on the vehicle side, improving user experience and satisfaction.

[0055] In this embodiment, a method for processing a voice broadcast on a vehicle terminal is provided, which can be used in a controller of a vehicle. Figure 3 is a flow chart of a method for processing a vehicle-side voice broadcast according to an embodiment of the present invention. Figure 3 As shown, the process includes the following steps:

[0056] Step S301, obtain the user's voice command, and determine the user's command intention according to the voice command. Figure 2 Step S201 of the illustrated embodiment will not be described in detail here.

[0057] Step S302: determine the domain type of the functional domain to which the voice command belongs according to the user's command intention, and determine a preset label of the domain type according to the functional domain label constraint rule.

[0058] Specifically, the above step S302 includes:

[0059] Step S3021, based on the user's command intention, select a preset application corresponding to the voice command from the vehicle function applications installed on the vehicle side.

[0060] Specifically, in an embodiment of the present invention, a large number of vehicle functions are deployed on the vehicle terminal, such as navigation function, music playback function, window control function, etc., to provide various services for users. In this case, after the vehicle terminal is controlled by voice commands to execute the corresponding vehicle function, if a voice reply can be made, the user's driving experience can be further improved. Therefore, an embodiment of the present invention screens all vehicle functions according to the user's command intention to determine whether the corresponding application is included. For example, according to the user's navigation intention, it is determined whether the vehicle function contains an application that can initiate navigation. It can be a navigation application on the vehicle terminal or a navigation application on a terminal device connected to the vehicle terminal, such as a mobile phone. However, if the corresponding application cannot be found, the vehicle terminal calls a fixed voice package in the local voice resource as a reply backup, such as replying "Hello, this service is temporarily unavailable to you". This is only an example and is not limited to this.

[0061] Step S3022: Determine the domain type of the functional domain to which the voice command belongs according to the preset application.

[0062] Specifically, in an embodiment of the present invention, the domain type of the functional domain to which the voice command falls is determined based on the found application, indicating that the vehicle terminal will provide the user with this type of functional service at this time. For example, according to the navigation application, the domain type of the functional domain to which the voice command belongs is determined to be the navigation functional domain. This is only an example and is not limited to this.

[0063] Step S3023, obtaining a preset functional domain label constraint rule, and determining a preset label of the domain type according to the functional domain label constraint rule.

[0064] Specifically, in an embodiment of the present invention, the functional domain label constraint rules are set with labels pre-labeled for different types of functional domains. Therefore, after determining the domain type of the functional domain to which the current voice command belongs, the corresponding preset labels can be directly determined according to the functional domain label constraint rules, including fixed reply labels and real-time reply labels, but not limited to this.

[0065] Step S303, based on the preset tags, determines whether it is necessary to call the language generation model pre-built in the cloud to perform voice broadcast according to the user's instruction intention.

[0066] Specifically, the above step S303 includes:

[0067] Step S3031: If the preset tag is a fixed reply tag, it is determined whether the local voice resource contains the corresponding first voice package according to the user instruction intention and the local voice resource. If it does, the first voice package is broadcasted.

[0068] Specifically, in the embodiment of the present invention, Figure 4As shown, the local voice resources contain voice packages that users often use in some fixed scenarios, such as voice packages for users when they go to work and get off work every day. When it is determined that the label corresponding to the domain type of the functional domain to which the user's current voice command belongs is a fixed reply label, the user's command intention and the local voice resource are directly matched with the text to determine whether the local voice resource contains the corresponding voice package. If it does, it is directly called and voice broadcast is performed. In this scenario, there is no need for voice generation through a large language model in the cloud, and the broadcast speed of the cockpit domain TTS (Text-to-Speech, speech synthesis / text to speech) reply can be achieved.

[0069] Step S3032: If not included, a resource call instruction is sent to the cloud, so that the cloud determines whether the cloud voice resources contain the corresponding second voice package according to the user's instruction intention and the cloud voice resources. If included, the second voice package sent by the cloud is broadcast and the second voice package is stored in the local voice resources.

[0070] Specifically, in an embodiment of the present invention, in the early stage of voice interaction between the user and the vehicle terminal, the number of voice packages stored in the local voice resources of the vehicle terminal is relatively small. In this case, if the voice package corresponding to the user's command intention cannot be matched in the local voice resource, the vehicle terminal sends a resource call instruction to the cloud, and the cloud performs text matching between the user's command intention and the local voice resource, and determines whether the cloud voice resource contains the corresponding voice package. If it does, the corresponding voice package is sent to the vehicle terminal, and the vehicle terminal plays the voice package and stores the voice package in the local voice resource. Therefore, the local voice resources are gradually enriched with the voice interaction between the user and the vehicle terminal.

[0071] Step S3033, if not included, sends a model call instruction to the cloud, so that the cloud generates a corresponding third voice package based on the user instruction intention and the language generation model, broadcasts the third voice package sent by the cloud, and stores the third voice package in the local voice resource.

[0072] Specifically, in the embodiment of the present invention, the cloud voice resources are also limited, so Figure 4 As shown in the figure, if the voice package corresponding to the user's command intent cannot be matched in the cloud voice resources, the vehicle side sends a model call instruction to the cloud, and the cloud calls the pre-built large language model to generate the corresponding voice package according to the user's command intent, and sends the corresponding voice package to the vehicle side, which plays the voice package and stores it in the local voice resources. At the same time, the cloud will also store the voice package in the server corresponding to the cloud voice resources. Therefore, both local voice resources and cloud voice resources will gradually become richer with the voice interaction between users and the vehicle side, and other users can also call the cloud voice resources.

[0073] The method for processing the voice broadcast on the vehicle side provided by the embodiment of the present invention determines the user's command intention according to the user's voice command, determines the domain type of the functional domain to which the voice command belongs according to the user's command intention, and determines the corresponding preset label according to the functional domain label constraint rule, and determines whether to call the language generation model pre-built in the cloud based on the preset label to perform voice broadcast according to the user's command intention. By constraining the functional domain with labels, the present invention can determine whether it is necessary to call the language generation model for voice broadcast according to the user's command intention, achieve efficient voice broadcast when no call is needed, and provide high-quality voice broadcast when call is needed, thereby providing users with efficient and high-quality voice broadcast on the vehicle side, improving user experience and satisfaction.

[0074] In this embodiment, a method for processing a voice broadcast on a vehicle terminal is provided, which can be used in a controller of a vehicle. Figure 5 is a flow chart of a method for processing a vehicle-side voice broadcast according to an embodiment of the present invention. Figure 5 As shown, the process includes the following steps:

[0075] Step S501, obtain the user's voice command, and determine the user's command intention according to the voice command. Figure 3 Step S301 of the illustrated embodiment will not be described in detail here.

[0076] Step S502: determine the domain type of the functional domain to which the voice command belongs according to the user's command intention, and determine the preset label of the domain type according to the functional domain label constraint rule. Figure 3 Step S302 of the illustrated embodiment will not be described in detail here.

[0077] Step S503, if the preset tag is a real-time reply tag, a model call instruction is sent to the cloud, so that the cloud generates a corresponding fourth voice package based on the user instruction intention and the language generation model, and broadcasts the fourth voice package sent by the cloud.

[0078] Specifically, in an embodiment of the present invention, if a user issues a voice instruction for question and answer or small talk, such as "What new movies are there recently?", the voice instruction is judged to be an inquiry intention, and the corresponding application cannot be filtered out from the vehicle computer functions because the determined label is a real-time reply label and the user hopes to get a real-time reply.

[0079] In some optional implementations, because the voice instructions such as user questions and answers or small talk are relatively flexible, the local voice resources cannot store the corresponding voice packets. Figure 4As shown, for the voice command of the real-time reply tag, the embodiment of the present invention needs to send a model call instruction to the cloud, that is, the cloud calls the large language model to generate a real-time voice package according to the user's command intention, and sends the corresponding voice package to the vehicle end, and the vehicle end plays the voice package. In this scenario, the vehicle end will not store the voice package, but only generate the corresponding reply language in real time to realize the question-and-answer communication with the user.

[0080] By performing label constraints on functional domains, the present invention can determine whether it is necessary to call a language generation model for voice broadcast according to the user's instruction intention, achieve efficient voice broadcast when no call is needed, and provide high-quality voice broadcast when call is needed, thereby providing users with efficient and high-quality voice broadcast on the vehicle side and improving user experience and satisfaction.

[0081] In this embodiment, a device for processing voice broadcasts on the vehicle side is also provided, which is used to implement the above-mentioned embodiments and preferred implementation modes, and will not be repeated hereafter. As used below, the term "module" can be a combination of software and / or hardware that implements a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0082] This embodiment provides a device for processing voice broadcasts at a vehicle terminal, such as Figure 6 As shown, including:

[0083] The intention determination module 601 is used to obtain the user's voice command and determine the user's instruction intention according to the voice command.

[0084] The label determination module 602 is used to determine the domain type of the functional domain to which the voice instruction belongs according to the user's instruction intention, and determine the preset label of the domain type according to the functional domain label constraint rule.

[0085] The voice broadcast module 603 is used to determine whether it is necessary to call the language generation model pre-built in the cloud based on the preset tags to perform voice broadcast according to the user's instruction intention.

[0086] In some optional implementations, the intention determination module 601 includes:

[0087] The speech recognition unit is used to perform speech recognition on the speech command to obtain speech text information.

[0088] The speech analysis unit is used to analyze the speech text information based on preset text matching rules to determine the user's instruction intention.

[0089] In some optional implementations, the tag determination module 602 includes:

[0090] The application screening unit is used to screen the preset applications corresponding to the voice commands from the vehicle function applications installed on the vehicle terminal based on the user's command intention.

[0091] The type determination unit is used to determine the domain type of the functional domain to which the voice instruction belongs according to a preset application.

[0092] The label determination unit is used to obtain a preset functional domain label constraint rule, and determine a preset label of the domain type according to the functional domain label constraint rule, wherein the preset label includes: a fixed reply label and a real-time reply label.

[0093] In some optional implementations, the voice announcement module 603 includes:

[0094] The local resource calling unit is used to determine whether the local voice resource contains the corresponding first voice package according to the user instruction intention and the local voice resource if the preset tag is a fixed reply tag, and if so, broadcast the first voice package.

[0095] The cloud resource calling unit is used to send a resource calling instruction to the cloud if it is not included, so that the cloud can determine whether the cloud voice resources contain the corresponding second voice package according to the user's instruction intention and the cloud voice resources. If it is included, the second voice package sent by the cloud is broadcast and the second voice package is stored in the local voice resources.

[0096] The first model calling unit is used to send a model calling instruction to the cloud if it is not included, so that the cloud generates a corresponding third voice package based on the user instruction intention and the language generation model, broadcasts the third voice package sent by the cloud, and stores the third voice package in the local voice resource.

[0097] The second model calling unit is used to send a model calling instruction to the cloud if the preset tag is a real-time reply tag, so that the cloud generates a corresponding fourth voice package based on the user's instruction intention and language generation model, and broadcasts the fourth voice package sent by the cloud.

[0098] The further functional description of each of the above modules and units is the same as that of the above corresponding embodiments and will not be repeated here.

[0099] The processing device for the vehicle-side voice broadcast in this embodiment is presented in the form of a functional unit, where the unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that executes one or more software or fixed programs, and / or other devices that can provide the above functions.

[0100] The embodiment of the present invention also provides a computer device having the above Figure 6The processing device for the voice broadcast on the vehicle side is shown.

[0101] See also Figure 7 , Figure 7 is a schematic diagram of the structure of a computer device provided by an optional embodiment of the present invention, such as Figure 7 As shown, the computer device includes: one or more processors 10, a memory 20, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. Various components are connected to each other using different buses for communication, and can be installed on a common mainboard or installed in other ways as needed. The processor can process the instructions executed in the computer device, including instructions stored in or on the memory to display the graphical information of the GUI on an external input / output device (such as, a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Similarly, multiple computer devices can be connected, and each device provides some necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system). Figure 7 A processor 10 is taken as an example.

[0102] The processor 10 may be a central processing unit, a network processor or a combination thereof. The processor 10 may further include a hardware chip. The hardware chip may be a dedicated integrated circuit, a programmable logic device or a combination thereof. The programmable logic device may be a complex programmable logic device, a field programmable gate array, a general purpose array logic or any combination thereof.

[0103] The memory 20 stores instructions executable by at least one processor 10, so that the at least one processor 10 executes the method shown in the above embodiment.

[0104] The memory 20 may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created according to the use of the computer device, etc. In addition, the memory 20 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some optional embodiments, the memory 20 may optionally include a memory remotely arranged relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0105] The memory 20 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk or a solid state drive; the memory 20 may also include a combination of the above types of memory.

[0106] The computer device further comprises a communication interface 30 for the computer device to communicate with other devices or a communication network.

[0107] The embodiment of the present invention also provides a computer-readable storage medium. The method according to the embodiment of the present invention can be implemented in hardware, firmware, or can be implemented as a computer code that can be recorded in a storage medium, or can be implemented as a computer code that is originally stored in a remote storage medium or a non-temporary machine-readable storage medium and will be stored in a local storage medium through a network download, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only storage memory, a random access memory, a flash memory, a hard disk or a solid-state hard disk, etc.; further, the storage medium can also include a combination of the above types of memories. It can be understood that a computer, a processor, a microprocessor controller, or programmable hardware includes a storage component that can store or receive software or computer code. When the software or computer code is accessed and executed by a computer, a processor, or hardware, the method shown in the above embodiment is implemented.

[0108] A part of the present invention may be applied as a computer program product, such as a computer program instruction, which, when executed by a computer, can call or provide the method and / or technical solution according to the present invention through the operation of the computer. Those skilled in the art should understand that the existence of the computer program instruction in a computer-readable medium includes, but is not limited to, a source file, an executable file, an installation package file, etc., and accordingly, the way in which the computer program instruction is executed by the computer includes, but is not limited to: the computer directly executes the instruction, or the computer compiles the instruction and then executes the corresponding compiled program, or the computer reads and executes the instruction, or the computer reads and installs the instruction and then executes the corresponding installed program. Here, the computer-readable medium may be any available computer-readable storage medium or communication medium accessible to the computer.

[0109] The embodiment of the present invention also provides a car, comprising: a storage unit for storing local voice resources; a controller connected to the storage unit for executing the above Figure 1 , Figure 2 , Figure 3 or Figure 5 The processing method of the vehicle-side voice broadcast is shown.

[0110] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations are all within the scope defined by the appended claims.

Claims

1. A method for processing voice broadcast at a vehicle terminal, characterized in that: The method comprises: Acquire a user's voice command and determine the user's command intention based on the voice command; Determine the domain type of the functional domain to which the voice command belongs according to the user command intention, and determine a preset label of the domain type according to the functional domain label constraint rule; Based on the preset tags, it is determined whether it is necessary to call a language generation model pre-built in the cloud to perform voice broadcast according to the user's instruction intention.

2. The method according to claim 1, characterized in that The determining the user instruction intention according to the voice instruction comprises: Performing voice recognition on the voice command to obtain voice text information; The voice text information is parsed based on preset text matching rules to determine the user instruction intention.

3. The method according to claim 1, characterized in that The determining the domain type of the functional domain to which the voice command belongs according to the user command intention, and determining the preset label of the domain type according to the functional domain label constraint rule, includes: Based on the user's command intention, select a preset application corresponding to the voice command from the vehicle function applications installed on the vehicle terminal; Determine the domain type of the functional domain to which the voice command belongs according to the preset application; A preset functional domain label constraint rule is obtained, and a preset label of the domain type is determined according to the functional domain label constraint rule, wherein the preset label includes: a fixed reply label and a real-time reply label.

4. The method according to claim 3, characterized in that The determining whether it is necessary to call a language generation model pre-built in the cloud based on the preset tag to perform voice broadcast according to the user instruction intention includes: If the preset tag is a fixed reply tag, judging whether the local voice resource contains the corresponding first voice package according to the user instruction intention and the local voice resource, and if so, broadcasting the first voice package; If not, a resource calling instruction is sent to the cloud, so that the cloud determines whether the cloud voice resource contains the corresponding second voice package according to the user instruction intention and the cloud voice resource; if it is contained, the second voice package sent by the cloud is broadcasted and the second voice package is stored in the local voice resource; If not included, a model call instruction is sent to the cloud, so that the cloud generates a corresponding third voice package based on the user instruction intention and the language generation model, and the third voice package sent by the cloud is broadcast, and the third voice package is stored in the local voice resource.

5. The method according to claim 3, characterized in that: The determining whether it is necessary to call a language generation model pre-built in the cloud based on the preset tag to perform voice broadcast according to the user instruction intention also includes: If the preset tag is a real-time reply tag, a model call instruction is sent to the cloud, so that the cloud generates a corresponding fourth voice package based on the user instruction intention and the language generation model, and broadcasts the fourth voice package sent by the cloud.

6. A vehicle-side voice broadcast processing device, characterized in that: The device comprises: An intention determination module, used to obtain a user's voice command and determine the user's instruction intention based on the voice command; A label determination module, used to determine the domain type of the functional domain to which the voice instruction belongs according to the user instruction intention, and determine a preset label of the domain type according to the functional domain label constraint rule; The voice broadcast module is used to determine whether it is necessary to call the language generation model pre-built in the cloud based on the preset tags to perform voice broadcast according to the user's instruction intention.

7. A computer device, characterized in that: include: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the method for processing the vehicle-side voice broadcast according to any one of claims 1 to 5 by executing the computer instructions.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute the method for processing the vehicle-side voice broadcast according to any one of claims 1 to 5.

9. A computer program product, characterized in that It includes computer instructions, and the computer instructions are used to enable a computer to execute the method for processing the vehicle-side voice broadcast according to any one of claims 1 to 5.

10. An automobile, characterized in that: include: A storage unit, used for storing local voice resources; A controller is connected to the storage unit and is used to execute the vehicle-side voice broadcast processing method according to any one of claims 1 to 5.

Citation Information

Cited By

  • Map voice broadcasting method, device and equipment based on large model and storage medium

    CN120403698A

  • Map voice broadcast method, device, equipment and storage medium based on large model

    CN120403698B