Vehicle voice interaction method, device, equipment and medium

Abstract

The invention discloses a vehicle voice interaction method, apparatus and device, and a medium. The method comprises the steps of obtaining dialogue information of a driver and passengers in a target vehicle; inputting the dialogue information into a local large model deployed in the target vehicle, and analyzing and processing the dialogue information through the local large model to obtain standard question information; wherein the local large model is an offline large model, and the standard question information comprises at least one of optimization question information, a first prompt word and answer form information; sending the standard question information to a cloud big model for analysis and processing to obtain first reply information fed back by the cloud big model; and playing the first reply information based on the voice dialogue function of the target vehicle. According to the technical scheme, through cooperative processing of the local large model and the cloud large model, the accuracy, safety and user experience of vehicle voice interaction are effectively improved. The method can be widely applied to the technical field of vehicles.

Description

Technical Field

[0001] The present application relates to the field of vehicle technology, and in particular to vehicle voice interaction methods, devices, equipment, and media. Background Art

[0002] In vehicle voice interaction scenarios, the Big Language Model serves as a core technology, leveraging natural language processing to understand driver and passenger voice commands and casual conversations and provide intelligent feedback. For example, it can answer navigation queries, recommend nearby services, adjust in-car devices, and even conduct multiple rounds of conversation to enhance the interactive experience. Leveraging cloud computing power, the Big Language Model can cover a wide range of semantic scenarios, from practical functions to entertaining chats. Its anthropomorphic expression also makes interactions more natural, making it a crucial addition to the smart cockpit's capabilities.

[0003] However, the practical application of large language models in vehicle voice interaction still faces some challenges. Due to the inherent complexity of the technology, it can be difficult for ordinary users to master the optimal interaction method, resulting in a high barrier to entry. The information input into large language models is relatively complex and has a variety of formats. The initial response provided by the model sometimes fails to accurately match actual needs, and may contain incomplete or inaccurate information. This reduces the practicality and reliability of voice interaction, impacting the user experience.

[0004] In summary, the technical problems existing in the relevant technologies need to be solved urgently. Summary of the Invention

[0005] The purpose of this application is to solve one of the technical problems existing in the related art to at least a certain extent.

[0006] To this end, one purpose of an embodiment of the present application is to provide a vehicle voice interaction method.

[0007] Another object of an embodiment of the present application is to provide a vehicle voice interaction device.

[0008] In order to achieve the above technical objectives, the technical solutions adopted in the embodiments of the present application include:

[0009] In a first aspect, an embodiment of the present application provides a vehicle voice interaction method, comprising:

[0010] Obtain conversation information of the driver and passengers in the target vehicle;

[0011] Inputting the conversation information into a local large model deployed in the target vehicle, analyzing and processing the conversation information by the local large model to obtain standard question information; wherein the local large model is an offline large model, and the standard question information includes at least one of optimization question information, a first prompt, and answer format information;

[0012] Sending the standard question information to the cloud-based large model for analysis and processing, and obtaining first reply information fed back by the cloud-based large model;

[0013] The first reply information is played based on the voice dialogue function of the target vehicle.

[0014] In addition, the vehicle voice interaction method according to the above embodiment of the present application may also have the following additional technical features:

[0015] Furthermore, in one embodiment of the present application, obtaining conversation information of the driver and passengers in the target vehicle includes:

[0016] Responding to an interaction instruction of the driver and passenger, acquiring audio data of a conversation of the driver and passenger;

[0017] Perform speech recognition on the conversation audio data to obtain the conversation information.

[0018] Furthermore, in one embodiment of the present application, after obtaining the conversation audio data of the driver and passenger, the method further includes:

[0019] The conversation audio data is subjected to denoising processing.

[0020] Furthermore, in one embodiment of the present application, the number of parameters of the local large model is smaller than the number of parameters of the cloud large model; the method further includes:

[0021] Detecting the network connection status with the cloud-based large model;

[0022] If the network connection state is disconnected, inputting the standard question information into the local large model for analysis and processing to obtain second reply information fed back by the local large model;

[0023] The second reply information is played based on the voice dialogue function of the target vehicle.

[0024] Furthermore, in one embodiment of the present application, the dialog information is input into a local large model deployed in the target vehicle, and the dialog information is analyzed and processed by the local large model to obtain standard question information, including:

[0025] Obtaining a preset second prompt; wherein the second prompt is used to instruct the local large model to analyze and process the input content to obtain standard question information;

[0026] The dialogue information and the second prompt are input into the local large model, and the dialogue information is analyzed and processed by the local large model to obtain standard question information.

[0027] Furthermore, in one embodiment of the present application, inputting the dialogue information and the second prompt into the local macro model includes:

[0028] Parsing the conversation information to determine original question information contained in the conversation information and context information corresponding to the original question information;

[0029] searching for associated historical information from local data of the target vehicle based on the context information;

[0030] The original question information, the context information, the history information and the second prompt are input into the local big model.

[0031] In a second aspect, an embodiment of the present application provides a vehicle voice interaction device, comprising:

[0032] an acquisition unit, configured to acquire conversation information of the driver and passengers in the target vehicle;

[0033] an input unit, configured to input the dialogue information into a local large model deployed in the target vehicle, and analyze and process the dialogue information by the local large model to obtain standard question information; wherein the local large model is an offline large model, and the standard question information includes at least one of optimization question information, a first prompt, and answer format information;

[0034] An analysis unit is configured to send the standard question information to a cloud-based large model for analysis and processing, and obtain first reply information fed back by the cloud-based large model;

[0035] A playing unit is used to play the first reply information based on the voice dialogue function of the target vehicle.

[0036] Furthermore, in one embodiment of the present application, the acquiring unit is specifically configured to:

[0037] Responding to an interaction instruction of the driver and passenger, acquiring audio data of a conversation of the driver and passenger;

[0038] Perform speech recognition on the conversation audio data to obtain the conversation information.

[0039] In a third aspect, an embodiment of the present application provides a computer device, including:

[0040] at least one processor;

[0041] at least one memory for storing at least one program;

[0042] When the at least one program is executed by the at least one processor, the at least one processor implements the vehicle voice interaction method described in the first aspect.

[0043] In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium storing a program executable by a processor, wherein the program executable by the processor is used to implement the vehicle voice interaction method described in the first aspect when executed by the processor.

[0044] The advantages and benefits of this application will be partially given in the following description, and partially become apparent from the following description, or learned through practice of this application:

[0045] In an embodiment of the present application, a vehicle voice interaction method is provided, which obtains conversation information of a driver and passenger in a target vehicle; inputs the conversation information into a local large model deployed in the target vehicle, and analyzes and processes the conversation information through the local large model to obtain standard question information; wherein the local large model is an offline large model, and the standard question information includes at least one of optimization question information, a first prompt, and answer form information; sends the standard question information to a cloud large model for analysis and processing, and obtains first reply information fed back by the cloud large model; and plays the first reply information based on the voice conversation function of the target vehicle. The technical solution of the present application effectively improves the accuracy, security, and user experience of vehicle voice interaction through the collaborative processing of the local large model and the cloud large model. The local large model pre-processes the original conversation information and generates structured standard question information (such as optimization question, prompt, or answer form), reducing the interference of fuzzy input on the cloud large model, thereby improving the accuracy of the reply and reducing information bias. At the same time, local processing avoids directly uploading sensitive conversation information to the cloud, enhancing privacy protection. The offline deployed local large model can also provide basic services when the network is poor, ensuring the stability of interaction. This solution lowers the threshold for using large language models while taking into account data security and response efficiency, making in-vehicle voice interaction more reliable and practical. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following introduction is made to the drawings of the embodiments of the present application or the related technical solutions in the prior art. It should be understood that the drawings introduced below are only for the convenience of clearly describing some embodiments of the technical solutions of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without any creative work.

[0047] Figure 1 A schematic diagram of an implementation environment of a vehicle voice interaction method provided in an embodiment of the present application;

[0048] Figure 2 A schematic diagram of a vehicle voice interaction method provided in an embodiment of the present application;

[0049] Figure 3 This is a structural diagram of a vehicle voice interaction device provided in an embodiment of the present application;

[0050] Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0051] The embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present application and are not to be construed as limiting the present application. For the step numbers in the following embodiments, they are provided only for the convenience of explanation and are not intended to limit the order of the steps. The order of execution of the steps in the embodiments can be adaptively adjusted according to the understanding of those skilled in the art.

[0052] In vehicle voice interaction scenarios, the Big Language Model serves as a core technology, leveraging natural language processing to understand driver and passenger voice commands and casual conversations and provide intelligent feedback. For example, it can answer navigation queries, recommend nearby services, adjust in-car devices, and even conduct multiple rounds of conversation to enhance the interactive experience. Leveraging cloud computing power, the Big Language Model can cover a wide range of semantic scenarios, from practical functions to entertaining chats. Its anthropomorphic expression also makes interactions more natural, making it a crucial addition to the smart cockpit's capabilities.

[0053] However, the practical application of large language models in vehicle voice interaction still faces some challenges. Due to the inherent complexity of the technology, it can be difficult for ordinary users to master the optimal interaction method, resulting in a high barrier to entry. The information input into large language models is relatively complex and has a variety of formats. The initial response provided by the model sometimes fails to accurately match actual needs, and may contain incomplete or inaccurate information. This reduces the practicality and reliability of voice interaction, impacting the user experience.

[0054] Moreover, since the conversation information in the vehicle environment is directly input into the large language model in the background for processing, there is a risk of privacy leakage and insufficient data security.

[0055] In view of this, an embodiment of the present application provides a vehicle voice interaction method, which obtains conversation information of a driver and passenger in a target vehicle; inputs the conversation information into a local large model deployed in the target vehicle, and analyzes and processes the conversation information through the local large model to obtain standard question information; wherein the local large model is an offline large model, and the standard question information includes at least one of optimization question information, a first prompt, and answer form information; sends the standard question information to a cloud large model for analysis and processing, and obtains a first reply information fed back by the cloud large model; and plays the first reply information based on the voice conversation function of the target vehicle. The technical solution of the present application effectively improves the accuracy, security, and user experience of vehicle voice interaction through the collaborative processing of the local large model and the cloud large model. The local large model preprocesses the original conversation information and generates structured standard question information (such as optimization question, prompt, or answer form), reducing the interference of fuzzy input on the cloud large model, thereby improving the accuracy of the reply and reducing information bias. At the same time, local processing avoids directly uploading sensitive conversation information to the cloud, enhancing privacy protection. The offline deployed local large model can also provide basic services when the network is poor, ensuring the stability of interaction. This solution lowers the threshold for using large language models while taking into account data security and response efficiency, making in-vehicle voice interaction more reliable and practical.

[0056] The technical solution of the present application has a wide range of applicability and can be adapted to vehicles of different brands and different architectures. Specifically, the present application achieves a high degree of adaptability to vehicles of different brands and different architectures through the collaborative architecture design of the local large model and the cloud large model: the local large model serves as an intermediate processing layer, which can flexibly connect to the voice input interface of various types of vehicle-mounted systems, and uniformly convert the original dialogue information of different formats into standard question information, shielding the differences in the underlying vehicle platforms. This standardized processing means that the cloud-based large model does not need to be customized and developed for specific models, and only needs to receive input in a unified format to generate high-quality first reply information. This layered decoupling design concept allows the solution to be quickly deployed on various models by simply lightweight adaptation of the local large model, greatly improving the universality and portability of the technology, and providing a standardized solution for cross-brand and cross-platform intelligent voice interaction.

[0057] Please refer to Figure 1 , Figure 1 The following is a schematic diagram of an implementation environment for a vehicle voice interaction method provided in an embodiment of the present application. In this implementation environment, the main hardware and software components involved include a terminal device 110 and a backend server 120. The terminal device 110 and the backend server 120 are in communication with each other.

[0058] Specifically, the vehicle voice interaction method provided in the embodiment of the present application can be executed based on the data interaction between the terminal device 110 and the background server 120. The terminal device 110 can be a vehicle-mounted terminal; the background server 120 can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN (Content Delivery Network), and big data and artificial intelligence platforms. On the terminal device 110 side, a local large model can be deployed, and on the background server 120 side, a cloud-based large model can be deployed. The two work together to provide accurate and reliable vehicle voice interaction services.

[0059] The terminal device 110 and the backend server 120 may establish a communication connection via a wireless network or a wired network. The wireless network or wired network uses standard communication technologies and / or protocols, and the network may be the Internet or any other network, including but not limited to a local area network (LAN), a metropolitan area network (MAN), a wide area network (WAN), a mobile, wired or wireless network, a private network or any combination of a virtual private network.

[0060] Of course, it is understandable that Figure 1 The implementation environment is only some optional application scenarios of the vehicle voice interaction method provided in the embodiment of this application. The actual application is not fixed. Figure 1 The hardware and software environment shown.

[0061] Below, in combination with the introduction of the aforementioned implementation environment, a vehicle voice interaction method provided in an embodiment of the present application is introduced and explained.

[0062] Please refer to Figure 2 , Figure 2 Schematic diagram of a vehicle voice interaction method provided by an embodiment of the present application, which includes but is not limited to:

[0063] Step 210: Acquire conversation information of the driver and passengers in the target vehicle;

[0064] Step 220: Input the conversation information into a local large model deployed in the target vehicle, and analyze and process the conversation information using the local large model to obtain standard question information; wherein the local large model is an offline large model, and the standard question information includes at least one of optimized question information, a first prompt, and answer format information;

[0065] Step 230: Send the standard question information to the cloud-based large model for analysis and processing, and obtain first reply information fed back by the cloud-based large model;

[0066] Step 240: Play the first reply information based on the voice dialogue function of the target vehicle.

[0067] In an embodiment of the present application, a vehicle voice interaction method is provided, which effectively improves the accuracy, security and user experience of vehicle voice interaction through the collaborative processing of a local large model and a cloud large model. The local large model pre-processes the original dialogue information, generates structured standard question information (such as optimized questions, prompts or answer forms), reduces the interference of fuzzy input on the cloud large model, thereby improving the accuracy of the answer and reducing information bias. At the same time, local processing avoids directly uploading sensitive dialogue information to the cloud, thereby enhancing privacy protection. The offline deployed local large model can also provide basic services when the network is poor to ensure interaction stability. While lowering the threshold for using large language models, this method takes into account data security and response efficiency, making in-vehicle voice interaction more reliable and practical.

[0068] In the embodiment of the present application, the vehicle to which the vehicle voice interaction method is applied can be recorded as the target vehicle. When executing the vehicle voice interaction method provided in the embodiment of the present application, the conversation information of the driver and passengers in the target vehicle can be obtained first. For example, when the driver and passengers need to use the vehicle voice interaction function provided by the large language model, they can trigger the relevant interaction instructions. For example, the driver and passengers can trigger the voice interaction function through a natural language wake-up word (such as "Hello XX") or a physical button. Then, the on-board microphone array in the target vehicle will collect voice data from multiple sound zones in the vehicle in real time to obtain the driver and passengers' conversation audio data. In the embodiment of the present application, the driver and passengers' conversation audio data can be either a conversation between the driver and the target vehicle or a conversation between different drivers and passengers, and this application does not impose any restrictions on this. After obtaining the driver and passengers' conversation audio data, it can be denoised first. For example, active noise reduction and sound source localization technology can be combined to effectively distinguish the driver and passenger voice inputs and filter background noise (such as engine sound, wind noise, etc.). The system performs endpoint detection on the original speech, automatically extracting valid conversation segments, and then converts the audio into text using automated speech recognition (ASR) technology. To improve recognition accuracy, the system dynamically adjusts pickup parameters based on vehicle status information (such as speed and window opening / closing). It also optimizes the acoustic model for the in-vehicle environment to accommodate different accents, speaking speeds, and common in-vehicle conversation scenarios (such as navigation queries, entertainment control, and vehicle settings).

[0069] In this embodiment of the present application, after obtaining conversation information, the conversation information can be input into a local large-scale model deployed in the target vehicle. The local large-scale model acts as a key intermediate layer to deeply structure the original conversation information. Specifically, after the system obtains the conversation information of the driver and passenger, the local large-scale model performs semantic parsing and intent recognition. Using contextual understanding technology, it converts any ambiguous expressions, colloquial expressions, or elliptical sentences into logically clear optimization problem information. For example, the local large-scale model can normalize the user's statement "I'm hungry, what's nearby to eat?" into "Please recommend a Chinese restaurant with a rating of 4 stars or above within 3 kilometers of the vehicle's current location." At the same time, the local large-scale model automatically adds a first prompt as a contextual constraint. This first prompt is used to provide richer contextual information for subsequent cloud-based large-scale model processing. For example, if a user needs to recommend a restaurant, the corresponding first prompt might be "The user is driving a new energy vehicle and is currently located in XXX during dinner time on a weekday. You are an experienced restaurant blogger. Please recommend a suitable restaurant based on the question information." If the user needs to recommend a tourist destination, the corresponding first prompt might be "You are a travel expert who specializes in introducing famous places of interest." Of course, it is understandable that the first prompt may vary depending on the application scenario, and this application does not limit this.

[0070] Furthermore, in this embodiment of the present application, the local large model intelligently determines the required answer format and generates corresponding answer format information. For example, if a user requests a restaurant recommendation, the answer format information may include "Please list three options in Chinese, including distance, signature dishes, and parking information." This application does not impose any restrictions on this.

[0071] As you can understand, in this embodiment, the original conversation information is processed by a local large model to generate standardized question information. This not only retains the convenience of natural interaction, but also significantly improves the accuracy and efficiency of subsequent cloud-based large model processing through standardization. At the same time, all sensitive information is desensitized on the vehicle side to ensure privacy and security.

[0072] In an embodiment of the present application, after the standard question information is transmitted to the cloud-based large model, the cloud-based large model can, based on its powerful computing power and massive knowledge base, parse the vehicle environment characteristics (such as new energy vehicle models, driving time, etc.) implicit in the standard question information, and combine the real-time updated geographic information data and service data to generate a customized reply that conforms to the vehicle-mounted scenario. In an embodiment of the present application, it is recorded as the first reply information. For example, for a restaurant recommendation request, the cloud-based large model will not only screen target locations that meet the conditions, but will also actively supplement vehicle-specific information such as the occupancy status of electric vehicle charging piles, the current road conditions, and the estimated arrival time. All reply content will undergo compliance verification of the vehicle-mounted scenario, automatically filter out content that is not suitable for driving scenarios (such as long video recommendations), and convert it into a colloquial expression that conforms to the characteristics of the in-vehicle voice broadcast through natural language generation technology. The cloud-based large model feeds back the first reply information to the target vehicle, and the target vehicle can play the first reply information based on the local voice dialogue function, thereby realizing vehicle voice interaction.

[0073] It is understandable that an embodiment of the present application provides a vehicle voice interaction method, which obtains conversation information of the driver and passengers in the target vehicle; inputs the conversation information into a local large model deployed in the target vehicle, and analyzes and processes the conversation information through the local large model to obtain standard question information; wherein, the local large model is an offline large model, and the standard question information includes at least one of optimized question information, a first prompt, and answer form information; sends the standard question information to the cloud large model for analysis and processing, and obtains first reply information fed back by the cloud large model; and plays the first reply information based on the voice conversation function of the target vehicle. The technical solution of the present application effectively improves the accuracy, security, and user experience of vehicle voice interaction through the collaborative processing of the local large model and the cloud large model.

[0074] Specifically, in some embodiments, the number of parameters of the local large model is smaller than the number of parameters of the cloud large model; the method further includes:

[0075] Detecting the network connection status with the cloud-based large model;

[0076] If the network connection state is disconnected, inputting the standard question information into the local large model for analysis and processing to obtain second reply information fed back by the local large model;

[0077] The second reply information is played based on the voice dialogue function of the target vehicle.

[0078] In the embodiment of the present application, the local large model can serve as a disaster recovery backup for the cloud-based large model, ensuring the continuity and reliability of the vehicle voice interaction service when the network connection status of the cloud-based large model is abnormal. Specifically, due to the limitations of the hardware equipment in the vehicle environment, the local large model can be a miniaturized large language model, and its parameter count can be much smaller than the cloud-based large model, for example, only 10%-20% of the cloud-based large model. It is easy to understand that due to the small number of parameters, the performance of the local large model will not be as good as the cloud-based large model, but it can retain basic semantic understanding capabilities.

[0079] When providing voice interaction services to drivers and passengers, the network connection status of the cloud-based large model can be continuously monitored. When a network interruption of the cloud-based large model is detected, it can automatically switch to local processing mode. At this time, the standard question information is no longer uploaded to the cloud, but the local large model directly generates the reply information, which is recorded as the second reply information in the embodiment of the present application. The second reply information can also be played based on the voice dialogue function of the target vehicle. For example, in a network-free environment such as a tunnel, the local large model can still process common requests based on the built-in offline knowledge base. This seamless switching mechanism ensures that basic functions are always available, and at the same time, the streamlined design of the local large model avoids response delays caused by complex calculations.

[0080] The step of inputting the dialogue information into a local large model deployed in the target vehicle and analyzing and processing the dialogue information by the local large model to obtain standard question information includes:

[0081] Specifically, in some embodiments, inputting the dialogue information into a local large model deployed in the target vehicle, and analyzing and processing the dialogue information by the local large model to obtain standard question information includes:

[0082] Obtaining a preset second prompt; wherein the second prompt is used to instruct the local large model to analyze and process the input content to obtain standard question information;

[0083] The dialogue information and the second prompt are input into the local large model, the dialogue information is analyzed and processed by the local large model, and standard question information is obtained.

[0084] In the embodiments of the present application, for the local large model, the intelligent processing and standard conversion of the original dialogue information can be realized by introducing the second prompt mechanism. The second prompt can be used to instruct the local large model to analyze and process the input content to obtain standard question information. For example, the second prompt can be "please integrate the input information, output standard question information, the standard question information includes question information, a prompt for instructing the cloud large model to reply to the question information, and answer form information", and the present application does not make any limitation.

[0085] The dialogue information and the second prompt are input into the local large model, the dialogue information is analyzed and processed by the local large model, and standard question information is obtained.

[0086] Specifically, in some embodiments, the inputting the dialogue information and the second prompt into the local large model comprises:

[0087] The dialogue information is parsed to determine the original question information contained in the dialogue information and the context information corresponding to the original question information;

[0088] According to the context information, the associated historical information is searched from the local data of the target vehicle;

[0089] The original question information, the context information, the historical information and the second prompt are input into the local large model.

[0090] In an embodiment of the present application, the local big model can achieve a deep understanding and precise analysis of the user's intentions through a multi-dimensional context fusion mechanism. Specifically, the original dialogue information can first be semantically deconstructed to identify the core question information (i.e., the original question information) and its associated context information. For example, when the user asks "Is there any room at that restaurant last time?", the local big model will parse out the core demand of "restaurant reservation" and extract the time dimension of "last time" and the reference relationship of "that restaurant" as key context information. Subsequently, the system will intelligently retrieve the historical interaction records, navigation destinations, favorite locations and other data stored locally in the vehicle, convert the fragmented context into specific parameters, and search for associated historical information. Then, the original question information, context information, historical information and the second prompt can be input into the local big model. The local big model uses multi-source information fusion technology to intelligently integrate the original question information, context information, historical data and the second prompt to form a complete semantic understanding framework, so that more accurate and reliable standard question information can be output.

[0091] It can be understood that in the embodiments of the present application, by integrating historical information, the problem of missing information in oral communication can be solved, and the scenario adaptability of questions can be greatly improved through the vehicle's unique local data (such as frequently visited places, usage habits); all data processing is completed on the vehicle side, which can fully protect the user's data security.

[0092] Reference Figure 3 In an embodiment of the present application, a vehicle voice interaction device is further provided, the device comprising:

[0093] An acquisition unit 310 is used to acquire conversation information of the driver and passengers in the target vehicle;

[0094] An input unit 320 is configured to input the conversation information into a local large model deployed in the target vehicle, and analyze and process the conversation information by the local large model to obtain standard question information; wherein the local large model is an offline large model, and the standard question information includes at least one of optimization question information, a first prompt, and answer format information;

[0095] An analysis unit 330 is configured to send the standard question information to a cloud-based large model for analysis and processing, and obtain first response information fed back by the cloud-based large model;

[0096] The playing unit 340 is configured to play the first reply information based on the voice dialogue function of the target vehicle.

[0097] It can be understood that the contents in the above method embodiments are applicable to the device embodiments, the device embodiments specifically implement the functions same as those of the above method embodiments, and achieve the beneficial effects same as those of the above method embodiments.

[0098] With reference to Figure 4 The embodiment of the present application provides an electronic device, comprising:

[0099] at least one processor 410;

[0100] at least one memory 420, configured to store at least one program;

[0101] When the at least one program is executed by the at least one processor 410, the at least one processor 410 implements the vehicle voice interaction method described above.

[0102] Similarly, the contents in the above method embodiments are applicable to the electronic device embodiments, the electronic device embodiments specifically implement the functions same as those of the above method embodiments, and achieve the beneficial effects same as those of the above method embodiments.

[0103] The embodiment of the present application further provides a computer readable storage medium, wherein the computer readable storage medium stores a program executable by the processor 410, and the program executable by the processor 410 is used for executing the vehicle voice interaction method described above when the program is executed by the processor 410.

[0104] Similarly, the contents in the above method embodiments are applicable to the computer readable storage medium embodiments, the computer readable storage medium embodiments specifically implement the functions same as those of the above method embodiments, and achieve the beneficial effects same as those of the above method embodiments.

[0105] In some alternative embodiments, the functions / operations mentioned in the block diagram can not occur in the order mentioned in the operation diagram. For example, depending on the functions / operations involved, two blocks shown in succession can actually be executed substantially simultaneously or the blocks can sometimes be executed in reverse order. In addition, the embodiments presented and described in the flowcharts of the present application are provided by way of example, and the purpose is to provide a more comprehensive understanding of the technology. The disclosed method is not limited to the operations and logical flows presented herein. Alternative embodiments are contemplated in which the order of various operations is changed and in which sub-operations described as part of larger operations are independently executed.

[0106] In addition, although the present application is described in the context of functional modules, it should be understood that, unless otherwise stated, one or more of the functions and / or features may be integrated into a single physical device and / or software module, or one or more functions and / or features may be implemented in separate physical devices or software modules. It is also understood that a detailed discussion of the actual implementation of each module is not necessary for understanding the present application. More specifically, given the properties, functions, and internal relationships of the various functional modules in the devices disclosed herein, the actual implementation of the module will be understood within the routine skills of an engineer. Therefore, a person skilled in the art can implement the present application as set forth in the claims using ordinary techniques without undue experimentation. It is also understood that the specific concepts disclosed are merely illustrative and are not intended to limit the scope of the present application, which is determined by the full scope of the appended claims and their equivalents.

[0107] If the function is implemented in the form of 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 the present application, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the various embodiments of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0108] The logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing the logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device). For purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0109] More specific examples (a non-exhaustive list) of the computer-readable medium include the following: an electrical connection (electronic) having one or more wires, a portable computer diskette (magnetic), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium can also be paper or another suitable medium upon which the program is printed, as the program can be electronically captured, for example, via optical scanning of the paper or other medium, then compiled, interpreted, or otherwise processed in a suitable manner, if necessary, and then stored in a computer memory.

[0110] It should be understood that aspects of the application can be implemented in hardware, software, firmware or combinations thereof. In the above described embodiments, various steps or methods can be implemented in software or firmware that is stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any of the following technologies, known in the art, or combinations thereof, can be used: a discrete logic circuit having logic gates for implementing logic functions upon data signals, an application specific integrated circuit having appropriate combinational logic gates, a programmable gate array (PGA), a field programmable gate array (FPGA), and / or the like.

[0111] In the above description of the present specification, the description referring to the terms "one embodiment", "another embodiment", or "certain embodiments" or the like means that a specific feature, structure, material or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present specification. The illustrative expressions of the above terms do not necessarily refer to the same embodiment or example throughout the present specification. Also, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0112] Although the embodiments of the present application have been shown and described, it will be appreciated by those skilled in the art that changes, modifications, alternatives and variations to these embodiments can be made without departing from the principles and spirit of the application, the scope of which is defined in the appended claims and their equivalents.

[0113] The above is a specific description of the preferred embodiments of the present application, but the present application is not limited to the embodiments, and those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of the present application, and these equivalent modifications or substitutions are included in the scope defined by the claims of the present application.

Claims

1. A vehicle voice interaction method, characterized in that: include: Obtain conversation information of the driver and passengers in the target vehicle; Inputting the conversation information into a local large model deployed in the target vehicle, analyzing and processing the conversation information by the local large model to obtain standard question information; wherein the local large model is an offline large model, and the standard question information includes at least one of optimization question information, a first prompt, and answer format information; Sending the standard question information to the cloud-based large model for analysis and processing, and obtaining first reply information fed back by the cloud-based large model; The first reply information is played based on the voice dialogue function of the target vehicle.

2. The vehicle voice interaction method according to claim 1, characterized in that: The acquiring of conversation information of the driver and passengers in the target vehicle includes: Responding to an interaction instruction of the driver and passenger, acquiring audio data of a conversation of the driver and passenger; Perform speech recognition on the conversation audio data to obtain the conversation information.

3. The vehicle voice interaction method according to claim 2, characterized in that: After obtaining the audio data of the conversation between the driver and passenger, the method further includes: The conversation audio data is subjected to denoising processing.

4. The vehicle voice interaction method according to claim 1, characterized in that: The number of parameters of the local large model is smaller than the number of parameters of the cloud large model; the method further includes: Detecting the network connection status with the cloud-based large model; If the network connection state is disconnected, inputting the standard question information into the local large model for analysis and processing to obtain second reply information fed back by the local large model; The second reply information is played based on the voice dialogue function of the target vehicle.

5. The vehicle voice interaction method according to claim 1, characterized in that: The step of inputting the dialogue information into a local large model deployed in the target vehicle and analyzing and processing the dialogue information by the local large model to obtain standard question information includes: Obtaining a preset second prompt; wherein the second prompt is used to instruct the local large model to analyze and process the input content to obtain standard question information; The dialogue information and the second prompt are input into the local large model, and the dialogue information is analyzed and processed by the local large model to obtain standard question information.

6. The vehicle voice interaction method according to claim 5, characterized in that: The step of inputting the dialogue information and the second prompt into the local large model includes: Parsing the conversation information to determine original question information contained in the conversation information and context information corresponding to the original question information; searching for associated historical information from local data of the target vehicle based on the context information; The original question information, the context information, the history information and the second prompt are input into the local big model.

7. A vehicle voice interaction device, characterized in that: include: an acquisition unit, configured to acquire conversation information of the driver and passengers in the target vehicle; an input unit, configured to input the dialogue information into a local large model deployed in the target vehicle, and analyze and process the dialogue information by the local large model to obtain standard question information; wherein the local large model is an offline large model, and the standard question information includes at least one of optimization question information, a first prompt, and answer format information; An analysis unit is configured to send the standard question information to a cloud-based large model for analysis and processing, and obtain first reply information fed back by the cloud-based large model; A playing unit is used to play the first reply information based on the voice dialogue function of the target vehicle.

8. The vehicle voice interaction device according to claim 7, characterized in that: The acquisition unit is specifically configured to: Responding to an interaction instruction of the driver and passenger, acquiring audio data of a conversation of the driver and passenger; Perform speech recognition on the conversation audio data to obtain the conversation information.

9. A computer device, characterized in that: include: at least one processor; at least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor implements the vehicle voice interaction method as described in any one of claims 1 to 6.

10. A computer-readable storage medium storing a program executable by a processor, characterized in that: The processor-executable program is used to implement the vehicle voice interaction method as described in any one of claims 1 to 6 when executed by the processor.

Images