A voice interaction method and apparatus

Abstract

The present disclosure provides a method, system and device for voice interaction, comprising obtaining a voice instruction input by a user, identifying the voice instruction by an artificial intelligence large model to obtain a user intent; matching the user intent with registered intents of multiple services to obtain a matching result, and determining a target service according to the matching result, the target service being used to execute at least one instruction corresponding to the user intent; generating a scene identifier corresponding to the user intent under a preset condition and performing scene registration to obtain a registered scene, the registered scene comprising: an intent type corresponding to the scene, the scene identifier, and a corresponding relationship between the scene identifier and the target service. The method of the present disclosure introduces an artificial intelligence large model into full-time vehicle voice interaction, and distributes the user intent and generated content to a suitable execution party based on scene registration. The distribution of generated content of the large model is solved while meeting the full-time vehicle voice interaction.

Description

Technical Field

[0001] This disclosure relates to the field of data processing, and more particularly to a voice interaction method and apparatus. Background Technology

[0002] Current in-vehicle voice technology has evolved to enable real-time, multi-user interaction, allowing commands to be issued from any time and location (multi-voice zone) within the vehicle, and enabling multiple people to participate in the same conversation. In this real-time, multi-user interaction scenario, voice commands containing intent need to be routed to the appropriate execution side to complete the command. For example, the driver might say "Navigate to Tiananmen Square," the passenger might say "I want to listen to crosstalk," and the rear passenger might say "I want to watch a movie." After these commands are recognized, the corresponding execution side needs to be located and executed. However, current voice interaction methods, when receiving multiple commands with different intents, can only process a single command containing intent at a time. For instance, these technologies need to process the driver's "Navigate to Tiananmen Square" command before processing the passenger's "I want to listen to crosstalk" command. Processing multiple voice commands with different intents simultaneously will confuse the execution side, making it impossible to distribute multiple voice commands with different intents to the appropriate execution side. Summary of the Invention

[0003] This disclosure provides a voice interaction method, apparatus, electronic device, storage medium, and program product. According to a first aspect of this disclosure, a voice interaction method is provided, comprising: acquiring a voice command input by a user; recognizing the voice command using an artificial intelligence model to obtain a user intent; matching the user intent with registered intents of multiple services to obtain a matching result, and determining a target service based on the matching result, the target service being used to execute at least one instruction corresponding to the user intent; generating a scene identifier corresponding to the user intent under preset conditions and registering the scene to obtain a registered scene, the registered scene including: an intent type corresponding to the scene, a scene identifier, and a correspondence between the scene identifier and the target service; the preset conditions including intent feature data of a preset type for the user intent, the preset type including at least one of dialogue interaction and data acquisition; in response to receiving new intent feature data of a preset type, matching the new intent feature data of the preset type with the registered scene, and if a corresponding target registered scene is matched, distributing the new intent feature data of the preset type to the target service corresponding to the target registered scene for execution according to the scene identifier of the target registered scene.

[0004] In some embodiments, after obtaining a user-inputted voice command and recognizing the voice command using an artificial intelligence model to obtain the user intent, the method further includes: in response to recognizing a single-sentence voice command to obtain multiple user intents, outputting first prompt information, the first prompt information being used to guide the user to select at least one user intent from the multiple user intents; obtaining selection information input by the user, and determining a single user intent of the voice command from the multiple user intents based on the selection information.

[0005] In some embodiments, before matching the user intent with the registration intents of multiple services to obtain a matching result, the method further includes: obtaining the intent corresponding to the service based on the functions and business scenarios of the multiple services; registering at least one intent for each of the multiple services to obtain the registration intents of the multiple services.

[0006] In some embodiments, matching a user's intent with the registration intents of multiple services to obtain a matching result, and determining a target service based on the matching result, includes: matching the registration intents of multiple execution services with the user's intent; in response to the existence of two or more services whose registration intents match the user's intent, outputting a second prompt message, the second prompt message being used to guide the user to select at least one from the two or more services; obtaining the user's input response information, determining the corresponding service based on the response intent of the response information, and identifying the service as the target service.

[0007] In some embodiments, the registration scenario further includes scenario parameter information. Under preset conditions, a scenario identifier corresponding to the user intent is generated and the scenario is registered. The registration scenario includes: under preset conditions, obtaining the intent type and generating a scenario identifier corresponding to the user intent; establishing a correspondence between the scenario identifier and the target service; and determining the scenario parameter information based on the user intent through the target service. The scenario parameter information includes one or more of the following: the display device information corresponding to the scenario, the identifier that needs to receive generative artificial intelligence data, the participant whitelist, and the registration time.

[0008] In some embodiments, in response to receiving new intent feature data of a preset type of dialogue interaction, matching the new intent feature data of a preset type with the registration scenario includes: obtaining a new voice command input by the user, recognizing the new voice command through an artificial intelligence big data model to obtain new intent feature data of a preset type of dialogue interaction; matching the new intent feature data of a preset type of dialogue interaction with intent types corresponding to multiple registration scenarios to obtain intent matching results; and determining the target registration scenario from multiple registration scenarios based on the intent matching results.

[0009] In some embodiments, determining the target registration scenario from multiple registration scenarios based on the intent matching result includes: if the new preset type of intent feature data of dialogue interaction matches the intent type corresponding to two or more registration scenarios in the multiple registration scenarios, obtaining parameter information of the new user intent; matching the parameter information of the new preset type of intent feature data of dialogue interaction with the scenario parameter information of two or more registration scenarios to obtain parameter information matching result; and determining the target scenario from two or more registration scenarios based on the parameter information matching result.

[0010] In some embodiments, the data information includes generative artificial intelligence data. In response to receiving new preset type intent feature data, matching the new preset type intent feature data with the registration scenario includes: in response to the user intent being preset type intent feature data and the preset type being data acquisition, processing the user intent and the corresponding scenario identifier using a large artificial intelligence model to obtain generative artificial intelligence data, wherein the generative artificial intelligence data is bound to the scenario identifier corresponding to the user intent; matching the scenario identifier corresponding to the user intent with scenario identifiers of multiple registration scenarios to obtain identifier matching results; and determining the target registration scenario from the multiple registration scenarios based on the identifier matching results.

[0011] In some embodiments, the method further includes: determining the display device corresponding to the registration scenario based on scenario parameter information of the registration scenario; canceling the registration scenario under a preset cancellation condition, wherein the preset cancellation condition includes at least one of the following: not receiving new preset type intent feature data matching the registration scenario for a preset period of time, or the user turning off the display device corresponding to the registration scenario.

[0012] According to a second aspect of this disclosure, a voice interaction device is provided, comprising: a recognition unit for acquiring voice commands input by a user, recognizing the voice commands using an artificial intelligence model to obtain the user's intent; a matching unit for matching the user's intent with registered intents of multiple services to obtain a matching result, and determining a target service based on the matching result, the target service being used to execute at least one instruction corresponding to the user's intent; a registration unit for generating a scene identifier corresponding to the user's intent under preset conditions and registering the scene to obtain a registered scene, the registered scene including: an intent type corresponding to the scene, a scene identifier, and a correspondence between the scene identifier and the target service; the preset conditions including intent feature data of a preset type for the user's intent, the preset type including at least one of dialogue interaction and data acquisition; and a distribution unit for responding to receiving new intent feature data of a preset type, matching the new intent feature data of the preset type with the registered scene, and if a corresponding target registered scene is matched, distributing the new intent feature data of the preset type to the target service corresponding to the target registered scene for execution based on the scene identifier of the target registered scene.

[0013] According to a third aspect of this disclosure, an electronic device is provided, comprising: at least one processor; and a memory connected to the at least one processor for voice interaction; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the method of the first aspect described above.

[0014] According to a fourth aspect of this disclosure, a non-transitory computer-readable storage medium is provided storing computer instructions, wherein the computer instructions are used to cause a computer to perform the method of the first aspect described above.

[0015] According to a fifth aspect of this disclosure, a computer program product is provided, including a computer program that, when executed by a processor, implements the method as described in the first aspect above.

[0016] The voice interaction method provided in the embodiments of this disclosure includes: acquiring a user-inputted voice command; recognizing the voice command using an artificial intelligence big data model to obtain the user intent; matching the user intent with the registered intents of multiple services to obtain a matching result; determining a target service based on the matching result; the target service being used to execute at least one instruction corresponding to the user intent; generating a scene identifier corresponding to the user intent under preset conditions and registering the scene to obtain a registered scene; the registered scene includes: the intent type corresponding to the scene, the scene identifier, and the correspondence between the scene identifier and the target service; the preset conditions include intent feature data of a preset type for the user intent, the preset type including at least one of dialogue interaction and data acquisition; in response to receiving new intent feature data of a preset type, matching the new intent feature data of a preset type with the registered scene; if a corresponding target registered scene is matched, distributing the new intent feature data of a preset type to the target service corresponding to the target registered scene for execution according to the scene identifier of the target registered scene. The method of this disclosure introduces an artificial intelligence big data model into in-vehicle voice interaction in the whole vehicle and in all times, and distributes user intent and generative content to the appropriate executor based on scene registration, thus solving the problem of distributing generative content of the big data model while satisfying the voice intent distribution under the whole vehicle and all times interaction.

[0017] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this application, nor is it intended to limit the scope of this application. Other features of this application will become readily apparent from the following description. Attached Figure Description

[0018] The accompanying drawings are provided to better understand this solution and do not constitute a limitation of this disclosure. Wherein:

[0019] Figure 1 A flowchart illustrating a voice interaction method provided in an embodiment of this disclosure;

[0020] Figure 2 A flowchart illustrating a voice interaction method provided in an embodiment of this disclosure;

[0021] Figure 3 An example diagram of a voice interaction system provided in an embodiment of this disclosure;

[0022] Figure 4 A flowchart illustrating a voice interaction method provided in an embodiment of this disclosure;

[0023] Figure 5 A flowchart illustrating a voice interaction method provided in an embodiment of this disclosure;

[0024] Figure 6 This is a schematic diagram of the structure of a voice interaction device provided in an embodiment of the present disclosure;

[0025] Figure 7 A schematic block diagram of an example electronic device 600 provided for embodiments of this disclosure. Detailed Implementation

[0026] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0027] Current in-vehicle voice technology has evolved to the point where it enables real-time, multi-user interaction throughout the vehicle. This means that voice commands can be issued at any time and in any location (multi-voice zone) within the vehicle, and multiple people can participate in the same conversation. Voice command data containing intent needs to be sent to the appropriate execution side to complete the execution of the command. As large language models have shown significant advantages in fields such as intelligent question answering, introducing them into in-vehicle voice systems is an inevitable trend.

[0028] This disclosure proposes a voice interaction system and method that introduces generative artificial intelligence into in-vehicle voice, enabling the distribution of user intent under full-vehicle, all-time interaction while being compatible with the distribution of generative content from large language models.

[0029] The following describes in detail, with reference to the accompanying drawings, a voice interaction method, apparatus, electronic device, storage medium, and program product proposed in this disclosure.

[0030] Figure 1 This disclosure provides a voice interaction method applied to a voice interaction system, the execution entity of which can be a processor. The processor can be, for example, a processor corresponding to a mobile device such as a mobile phone, computer, or tablet; a processor corresponding to a smart wearable device such as smart glasses or smartwatches; or a processor of a vehicle, such as the main control chip mounted on the vehicle's infotainment system.

[0031] like Figure 1 As shown, the method includes the following steps:

[0032] Step 101: Obtain the user's voice input command, and use an artificial intelligence big data model to recognize the voice command and obtain the user's intent.

[0033] In some embodiments of this disclosure, large artificial intelligence models include Large Language Model (LLM), ChatGPT (Chat Generative Pre-trained Transformer), multimodal large models, and multimodal cognitive large models, etc.

[0034] In the embodiments of this disclosure, the large AI model used for speech recognition can handle a variety of natural language tasks, such as text classification, question answering, and dialogue, meeting the needs of voice interaction in multiple scenarios, such as intelligent customer service, smart homes, and autonomous driving.

[0035] In some embodiments, in addition to obtaining user intent through a large artificial intelligence model, for scenarios that require receiving multiple rounds of voice commands or dialogues, intelligent question answering can also be achieved by obtaining generative artificial intelligence data through a large artificial intelligence model. Generative artificial intelligence data includes images, text, audio, etc.

[0036] For example, in real-time voice interaction throughout the vehicle, if the front passenger says "I want to listen to crosstalk" or the rear passenger says "What's the difference between L8 and L9?", the AI ​​big data model can obtain and process the voice commands to determine the corresponding user intent, such as the intent to play crosstalk or the intent to have a conversation. For the front passenger's intent to play crosstalk, it generates a prompt message "Which one do you want to listen to?" to guide the user to select the specific playback target. For the rear passenger's request to "What's the difference between L8 and L9?", it generates corresponding dialogue information such as "The difference between L8 and L9 is..." to achieve intelligent question answering.

[0037] Step 102: Match the user intent with the registration intent of multiple services to obtain the matching result, and determine the target service based on the matching result. The target service is used to execute at least one instruction corresponding to the user intent.

[0038] In some embodiments, after the voice command is understood by the large-scale artificial intelligence model, it needs to find the corresponding service to execute. This is achieved by matching the user's intent with the registration intent of multiple services, and assigning the user's intent to the target service.

[0039] In some embodiments, in in-vehicle voice interaction scenarios, such as the driver saying "Navigate to Tiananmen Square," the passenger saying "I want to listen to crosstalk," and the rear passenger saying "The difference between L8 and L9," these voice commands are understood by the AI ​​big data model and then need to find the corresponding service to execute. For the driver, a navigation application is needed to execute the intention to navigate to Tiananmen Square; for the passenger's command, a media application is needed to execute the intention to play crosstalk; and for the rear passenger, a business processing unit is needed to receive the dialogue text or voice response generated by the AI ​​big data model and display it to the rear passenger.

[0040] Step 103: Generate a scene identifier corresponding to the user intent under preset conditions and register the scene to obtain a registered scene. The registered scene includes: the intent type corresponding to the scene, the scene identifier, and the correspondence between the scene identifier and the target service. The preset conditions include intent feature data of the user intent being of a preset type. The preset type includes at least one of dialogue interaction and data information acquisition.

[0041] In some embodiments, the dialogue interaction type includes continuously receiving voice commands from the user. For example, if the user says "I want to watch a movie", multiple movies are searched, and the user is prompted which one they want to watch, then the system needs to receive commands such as "first one" or "movie 'The Third Brigade'".

[0042] In some embodiments, the types of data information to be acquired include acquiring relevant data information based on the current intent feature data, such as acquiring the answer "The difference between L8 and L9 is..." based on the question "What does a Corgi look like?", or acquiring pictures of Corgis based on the intent "What does a Corgi look like?".

[0043] In some embodiments, after a user intent is distributed to a target service, during the execution of the user intent, it can be determined whether the user intent is intent feature data of a preset type by judging whether the execution of the user intent requires obtaining further user intent and whether the execution of the user intent requires receiving generative artificial intelligence data.

[0044] In some embodiments, the intent type corresponding to a scenario is determined based on the instruction intent expected to be received in that scenario.

[0045] For example, if a user says "I want to listen to crosstalk," the media service performs a crosstalk search, finds multiple results, and the user needs to choose one. The media service expects to receive instructions from the user such as "first" or "next page." In this case, the intent type corresponding to the registration scenario is a selection intent.

[0046] In some embodiments, the execution of a user intent requires obtaining further user intents. For example, when the execution of a user intent requires the user to make further choices, a scenario is registered for the user intent, a scenario identifier is generated, and the intent type corresponding to the scenario is defined.

[0047] For example, if the driver says "navigate to Tiananmen Square", the driver's intention to navigate to Tiananmen Square is assigned to the navigation service, and the navigation route selection interface is displayed. The navigation service needs to obtain further intents from the user, such as which route the user wants to choose. At this time, the system registers a scenario for the driver's intention to navigate to Tiananmen Square, generates a scenario identifier, and assigns the scenario to the intent type. The scenario identifier is then bound to the navigation service.

[0048] In some embodiments, when the execution of a user intent requires receiving generative artificial intelligence data, such as when the execution of a user intent requires obtaining dialogue information generated based on the user's voice command, registering a scenario for the user intent, generating a scenario identifier, specifying the intent type corresponding to the scenario as a dialogue intent, and identifying the need to receive generative artificial intelligence data, etc.

[0049] For example, when the person in the back says "the difference between L8 and L9", the dialogue intent of the person in the back is assigned to the dialogue service. At this time, it is necessary to receive the content generated based on the question from the person in the back, such as "the difference between L8 and L9 is...". At this time, a scenario is registered for this intent, and the scenario identifier is bound to the dialogue service.

[0050] Step 104: In response to receiving new preset type intent feature data, match the new preset type intent feature data with the registration scenario. If a matching target registration scenario is found, distribute the new preset type intent feature data to the target service execution corresponding to the target registration scenario according to the scenario identifier of the target registration scenario.

[0051] In some embodiments, scene matching is performed in response to receiving new preset type intention feature data of dialogue interaction and / or in response to receiving new preset type intention feature data of data acquisition, such as receiving new user intentions in multi-turn voice commands and / or acquiring generative artificial intelligence data.

[0052] In some embodiments, when a new user intent is received, it is matched with the registration scenario. If the new user intent is related to the registration scenario, it can be assigned to the target service corresponding to the registration scenario, thereby realizing the distribution of user intents throughout the vehicle and at all times.

[0053] For example, when executing the driver's intention "navigate to Tiananmen Square," the navigation displays multiple routes and asks the user which route to choose. When the passenger says "I want to watch a movie," the media displays multiple movies and asks the user which one to watch. Since both actions require obtaining further user intent, scenarios are registered for them separately, obtaining scenario 1 and scenario 2. If a new intent "I want to watch the movie 'Invincible'" is received at this time, the user intent is matched with scenario 1 and scenario 2. If the matching result shows that the user intent matches scenario 2 the most, then the user is assigned the media service corresponding to scenario 2.

[0054] In some embodiments, when the execution of a user intent requires receiving generative artificial intelligence data, a scenario is registered for the user intent, a scenario identifier is generated, and the user intent and the corresponding scenario identifier are simultaneously sent to an artificial intelligence big data model for processing to obtain generative artificial intelligence data. The generative artificial intelligence data is bound to the scenario identifier corresponding to the user intent. In response to receiving the generative artificial intelligence data, the corresponding registered scenario is matched according to the bound scenario identifier, and the generative artificial intelligence data is allocated to the target service corresponding to the registered scenario.

[0055] For example, to execute the user intent "the difference between L8 and L9", a dialogue question and answer is required. The dialogue service registers a scenario for the user intent, generates a scenario identifier, and sends "the difference between L8 and L9" and the corresponding scenario identifier to ChatGPT for processing, generating the content "the difference between L8 and L9 is...". This generated content is bound to the scenario identifier corresponding to "the difference between L8 and L9". Therefore, when this generated content is received, the corresponding registered scenario can be matched according to its bound scenario identifier, and it can be assigned to the dialogue service corresponding to the registered scenario.

[0056] In summary, according to the embodiments of this disclosure, the method includes acquiring a user-inputted voice command, recognizing the voice command using an artificial intelligence model to obtain the user intent; matching the user intent with the registered intents of multiple services to obtain a matching result, and determining a target service based on the matching result, the target service being used to execute at least one instruction corresponding to the user intent; generating a scene identifier corresponding to the user intent under preset conditions and registering the scene to obtain a registered scene, the registered scene including: the intent type corresponding to the scene, the scene identifier, and the correspondence between the scene identifier and the target service; the preset conditions include intent feature data of the user intent being of a preset type, the preset type including dialogue interaction, acquisition, etc. The method involves taking at least one type of data information; in response to receiving new user intent and / or generative artificial intelligence data, matching the new user intent and / or generative artificial intelligence data with the registration scenario; if a corresponding target registration scenario is matched, distributing the new user intent and / or generative artificial intelligence data to the target service execution corresponding to the target registration scenario according to the scenario identifier of the target registration scenario. The method disclosed herein introduces a large artificial intelligence model into voice interaction, and distributes user intent and generative artificial intelligence data to the appropriate executor based on scenario registration. While satisfying the voice intent distribution under full-vehicle, all-time interaction, it also solves the problem of distributing generative content of the large model.

[0057] based on Figure 1 The embodiment shown, Figure 2 This is a flowchart illustrating a voice interaction method provided in an embodiment of the present disclosure. This method can be applied to... Figure 3The voice interaction system shown is an example of one that can operate in a vehicle's infotainment system.

[0058] The method includes the following steps 201-2010.

[0059] Step 201: Obtain the voice command input by the user, and use the large artificial intelligence model to recognize the voice command to obtain the user's intent.

[0060] In some embodiments, such as Figure 3 In the voice interaction system shown, the AI ​​capability layer recognizes and processes the voice commands issued by the user to obtain the user's intent. The AI ​​capability layer provides AI capabilities including Natural Language Understanding (NLU), Natural Language Generation (NLG), and generative artificial intelligence.

[0061] In some embodiments, after obtaining a user-inputted voice command and recognizing the voice command using an artificial intelligence model to obtain the user intent, the method further includes: in response to recognizing a single-sentence voice command to obtain multiple user intents, outputting first prompt information, the first prompt information being used to guide the user to select at least one user intent from the multiple user intents; obtaining selection information input by the user, and determining a single user intent of the voice command from the multiple user intents based on the selection information.

[0062] In some embodiments, such as Figure 3 As shown, the dialogue management (hereinafter referred to as DM) of the voice interaction system performs semantic arbitration on multiple user intentions corresponding to the voice command to determine the individual user intention of the voice command. After determination, the intention is passed to the AI ​​assistant management (hereinafter referred to as CM), and the CM allocates it to the Copilot of the target service.

[0063] For example, if a user says "I want to see the sky," there could be multiple intentions, such as the intention to open the sunroof for vehicle control or the intention to search for a movie (i.e., the movie "Sky"). In this case, the DM can ask the user whether they want to open the sunroof or watch a movie. If it is determined that the user's intention is to open the car window, the DM can answer the question.

[0064] Step 202: Obtain the intent corresponding to the service based on the functions and business scenarios of multiple services.

[0065] Step 203: Register at least one intent for each of the multiple services to obtain the registration intents for the multiple services.

[0066] In some embodiments, such as Figure 3The voice interaction system shown includes multiple AI assistants (hereinafter referred to as Copilots), each Copilot corresponding to a service. For example, for the navigation service, there is a navigation Copilot that manages one or more navigation applications to execute the user's navigation-related intent, such as the intent to "navigate to Tiananmen Square". For the media service, there is a media Copilot that manages one or more media applications to execute the user's media-related intent, such as the intent to "play a movie". In addition, there are also vehicle control Copilots, general dialogue Copilots, etc.

[0067] In some embodiments, when the application starts, the corresponding intent is registered for the Copilot responsible for the application. For example, when the media application starts, the intent related to media search and playback control, such as "play music", "I want to watch a movie", "I want to listen to crosstalk", and "fast forward 30 seconds", is registered for the media Copilot. For the vehicle control Copilot, the intents such as "open the trunk", "turn on the reading lights", and "turn on the seat massage" are registered. Other Copilots are similar and will not be described in detail.

[0068] Step 204: Match the user intent with the registration intent of multiple services to obtain a matching result, and determine the target service based on the matching result. The target service is used to execute at least one instruction corresponding to the user intent.

[0069] In some embodiments, matching a user's intent with the registration intents of multiple services to obtain a matching result, and determining a target service based on the matching result, includes: matching the registration intents of multiple execution services with the user's intent; in response to the existence of two or more services whose registration intents match the user's intent, outputting a second prompt message, the second prompt message being used to guide the user to select at least one from the two or more services; obtaining the user's input response information, determining the corresponding service based on the response intent of the response information, and identifying the service as the target service.

[0070] In some embodiments, such as those applied to Figure 3 In the voice interaction system shown, when multiple Copilots register the same intent and the user triggers the intent, the human CM arbitrates the intent and asks the user which Copilot to use for execution.

[0071] For example, if a user says "I want to watch a movie," and iQiyi and Tencent Video each register their intent to search for video as a Copilot, then the CM can ask the user which application to use for the search. It should be noted that in other implementations, there may be a situation where one media Copilot manages all media applications, which is not limited in this disclosure.

[0072] Step 205: Generate a scene identifier corresponding to the user intent under preset conditions and register the scene to obtain a registered scene. The registered scene includes: the intent type corresponding to the scene, the scene identifier, and the correspondence between the scene identifier and the target service. The preset conditions include intent feature data of the user intent being of a preset type, and the preset type includes at least one of dialogue interaction and data information acquisition.

[0073] In some embodiments of this disclosure, after introducing generative artificial intelligence into the full-vehicle, all-time voice interaction, the distribution of user voice commands to the target service for execution includes three cases: first, the command is executed directly and the execution result is returned; second, the command requires further instructions from the user, including scenarios that need to support multi-turn dialogue; and third, it needs to receive AI-generated content, including scenarios that need to support intelligent question answering. In the latter two cases, the intent corresponding to the user's voice command is registered for the scenario.

[0074] In some embodiments of this disclosure, when it is necessary to receive further user intent and / or artificial intelligence-generated content during the execution of user intent, a scene identifier corresponding to the user intent is generated and the scene is registered.

[0075] In some embodiments, the registration scenario further includes scenario parameter information. Under preset conditions, a scenario identifier corresponding to the user intent is generated and the scenario is registered. The registration scenario includes: under preset conditions, obtaining the intent type and generating a scenario identifier corresponding to the user intent; establishing a correspondence between the scenario identifier and the target service; and determining the scenario parameter information based on the user intent through the target service. The scenario parameter information includes one or more of the following: the display device information corresponding to the scenario, the identifier that needs to receive generative artificial intelligence data, the participant whitelist, and the registration time.

[0076] In some embodiments, the registration process is often accompanied by a UI display, i.e., associated with a screen inside the vehicle.

[0077] In some embodiments, the scene ID generation rule is as follows: "sceneid"-ScreenId-timestamp-6-digit random number; each field is the identifier header, the screen corresponding to the scene, the timestamp, and the random number. In addition, other parameters can be defined to achieve specific purposes. For example, the following parameters can be defined: intent whitelist including the intent that Copilot wants to receive instructions at this time, screen information including the screen corresponding to the scene, participant whitelist including whether the scene is a multi-person scene, i.e., whether to receive instructions from other people, screen information including the screen corresponding to the scene, and parameters that indicate that the scene needs to receive generative human data, etc.

[0078] In some embodiments, applied to Figure 3The voice interaction system shown in the figure can register a scene and generate a scene ID when Copilot needs to dynamically receive intent or receive generative artificial intelligence data. The relationship between the scene ID and Copilot is recorded in CM and the scene information is transmitted to DM.

[0079] For example, a user says "I want to listen to crosstalk," and the media Copilot performs a crosstalk search, finding multiple results. The user needs to select one, and then Copilot displays the searched media resources and announces, "Multiple results found, which one do you want to listen to?" At this point, Copilot wants to receive instructions from the user such as "first" or "next page." Copilot will then register the scene, generating a scene ID such as B, and an intent whitelist including the intents that Copilot wants to receive, such as "first" or "next page." The relationship between scene B and media Copilot will be recorded in CM, and the scene information will be passed to DM.

[0080] For example, if a user says "compare L8 and L9", they will be assigned to the general dialogue Copilot. At this time, the user expects Copilot to receive the intelligent answer generated for the question. The general dialogue Copilot will register the scenario and generate a scenario ID such as C, which identifies the parameter information that the scenario needs to receive generative human data. The relationship between scenario C and general dialogue Copilot will be recorded in CM and the scenario information will be passed to DM.

[0081] Step 206: In response to receiving new preset type intent feature data, match the new preset type intent feature data with the registration scenario. If a matching target registration scenario is found, distribute the new preset type intent feature data to the target service execution corresponding to the target registration scenario according to the scenario identifier of the target registration scenario.

[0082] In some embodiments, scene matching is performed in response to receiving new intent feature data of a preset type of dialogue interaction and / or in response to receiving new intent feature data of a preset type of data acquisition, such as receiving new user intent in multi-turn voice commands and / or acquiring generative artificial intelligence data.

[0083] In some embodiments, such as Figure 3In the scenario shown, when Copilot registers a scenario, the CM is responsible for generating a scenario identifier. The relationship between the scenario identifier and Copilot is recorded in the CM, and the scenario information is passed to the DM. When there is a new user intent or AIGC related to the scenario, the DM is responsible for associating the new user intent or generative artificial intelligence data with the scenario ID and then passing it to the CM. The CM then routes the new user intent or generative artificial intelligence data to the Copilot that registered the scenario based on the scenario ID.

[0084] In some embodiments, such as Figure 4 As shown, in response to receiving new preset type intent feature data of dialogue interaction, matching the new preset type intent feature data with the registered scenario includes steps 301-303.

[0085] Step 301: Obtain the new voice command input by the user, and use the artificial intelligence big data model to recognize the new voice command to obtain new preset type of intention feature data for dialogue interaction.

[0086] Step 302: Match the new preset type of intent feature data (dialogue interaction) with the intent types corresponding to multiple registration scenarios to obtain intent matching results.

[0087] In some embodiments, when the driver says "Navigate to Tiananmen Square," the navigation Copilot displays a navigation route selection interface, prompting "Which route do you want to choose?" The registered scenario ID is A, and the intent list is the selected intent. When the passenger says "I want to watch a movie," the media Copilot pops up a power resource card, prompting "Multiple results found, which one do you want to watch?" and registers a scenario ID as B, with the intent list being the selected intent. If the driver says "The first one," the voice command is processed to obtain a new intent as the selected intent. Through intent matching, it can be seen that this intent matches the intent lists of registered scenario A and registered scenario B.

[0088] Step 303: Based on the intent matching results, determine the target registration scenario from multiple registration scenarios.

[0089] In some embodiments, determining a target registration scenario from multiple registration scenarios based on intent matching results includes: if new intent feature data of the preset type being dialogue interaction matches the intent types corresponding to two or more registration scenarios in the multiple registration scenarios, obtaining parameter information of the new intent feature data of the preset type being dialogue interaction; matching the parameter information of the new intent feature data of the preset type being dialogue interaction with the scenario parameter information of two or more registration scenarios to obtain parameter information matching results; and determining a target scenario from two or more registration scenarios based on the parameter information matching results.

[0090] In some embodiments, the parameter information of the new preset type of intention feature data for dialogue interaction includes the voice region or user corresponding to the source of the voice command, and the voice region includes the driver's seat, passenger seat, rear seat, etc.

[0091] In some embodiments, when the intent whitelists corresponding to multiple scenarios include the same intent and the user triggers the intent, the DM performs scenario arbitration on multiple scenarios, selects one scenario, associates the intent with the scenario identifier of the scenario and passes it to the CM, and then the CM sends it to the Copilot that registered the scenario.

[0092] In some embodiments, DM performs scene arbitration based on scene parameters, including scene arbitration based on screen information, a whitelist of participants, and one or more of the registration time.

[0093] In some embodiments, the whitelist of participants can be used to determine whether the registration scenario is a multi-person scenario. If it is a multi-person scenario, instructions from other people can be received. The whitelist of participants and the new preset type of intention feature data for dialogue interaction correspond to the voice region or user of the voice instruction source to determine whether the registration scenario matches.

[0094] For example, if the driver says "Navigate to Tiananmen Square," the navigation Copilot displays a route selection interface, prompting "Which route do you want to choose?" The registered scenario ID is A, and the intent list contains the selected intent. If the passenger says "I want to watch a movie," the media Copilot pops up a power resource card, and the TTS prompts "Multiple results found, which one do you want to watch?" and registers a scenario ID of B, with the intent list also containing the selected intent. If the passenger in the back says "The first one," based on the intent matching results, both registered scenarios A and B match. If registered scenario A is not a multi-person scenario, but registered scenario B is, then the selection intent of "The first one" is sent to the media Copilot corresponding to scenario B, and the operation of selecting to play the first movie is executed.

[0095] In some embodiments, screen information includes the screen associated with the scene and the audio zone where the screen is located. The audio zone includes the driver's seat, passenger seat, rear seat, etc. It is determined whether the audio zone from which the voice command corresponding to the new preset type of dialogue interaction intent feature data originates is the same as the audio zone where the screen associated with the registered scene is located. If they are the same audio zone, the registered scene is determined to be the target scene.

[0096] For example, the driver says, "Navigate to Tiananmen Square." The navigation Copilot displays a route selection interface, prompting "Which route do you want to choose?" The registered scenario ID is A, and the intent list is the selected intent. The rear passenger says, "I want to watch a movie." The media Copilot pops up a power resource card, and the TTS prompts, "Multiple results found, which one do you want to watch?" It registers a scenario ID, B, and the intent list is also the selected intent. At this point, the driver says, "The first one." Based on the intent matching results, both registered scenarios A and B match. If the screen associated with registered scenario A is in the driver's seat, and the screen associated with registered scenario B is in the rear passenger seat, then the selection intent of "the first one" is sent to the navigation Copilot corresponding to scenario A, executing the operation of selecting the first route and starting navigation, thus completing the instruction execution.

[0097] In some embodiments, if a new user intent matches the intent type corresponding to two or more registration scenarios among multiple registration scenarios, the registration scenario with the earlier registration time is determined as the target scenario based on the registration time of the registration scenario.

[0098] In some embodiments, such as Figure 5 As shown, the data information includes generative artificial intelligence data. In response to receiving new preset type intent feature data, the matching of the new preset type intent feature data with the registered scenario includes steps 401-403.

[0099] Step 401: In response to user intent being intent feature data of a preset type and the preset type being data acquisition, the user intent and the scene identifier corresponding to the user intent are processed through an artificial intelligence big data model to obtain generative artificial intelligence data, wherein the generative artificial intelligence data is bound to the scene identifier corresponding to the user intent.

[0100] Step 402: Match the scene identifier corresponding to the user intent with the scene identifiers of multiple registered scenes to obtain the identifier matching result.

[0101] Step 403: Determine the target registration scenario from multiple registration scenarios based on the identifier matching results.

[0102] In some embodiments, such as Figure 3In the voice interaction system shown, when someone in the back row says "compare L8 and L9," they need to receive generative artificial intelligence data. The general dialogue Copilot registers a scene ID of C for this intent, and sends the intent and the registered scene ID to the GTP for processing to generate comparison information between L8 and L9. The comparison information will be identified as AI data with scene ID C in the DM. If the currently registered scene includes scene A corresponding to the navigation Copilot, scene B corresponding to the media Copilot, and scene C corresponding to the general dialogue Copilot, then the AI ​​data will be distributed to the general dialogue Copilot by the CM according to the identified scene ID C.

[0103] Step 207: Under preset cancellation conditions, cancel the registration scenario. The preset cancellation conditions include at least one of the following: not receiving new preset type intent feature data matching the registration scenario for a preset period of time, or the user turning off the display device corresponding to the registration scenario.

[0104] In some embodiments, the display device corresponding to the registration scene is determined based on the scene parameter information of the registration scene.

[0105] In some embodiments, Copilot may deregister a scenario when it does not receive AI layer processing results, including new user intents and / or generative artificial intelligence data, within a preset time, or when the user closes the display device corresponding to the registered scenario. For example, if a user says "I want to watch a movie," the screen displays a UI showing the searched movie resources and asking the user which one they want to watch. If the user does not issue any further instructions for a long time, the UI disappears, or the user closes the UI directly. In this case, Copilot deregisters the scenario.

[0106] In summary, according to the embodiments of this disclosure, AI capabilities are distributed based on scenarios, enabling real-time, multi-user interaction of in-vehicle voice commands while supporting generative content distribution of large models.

[0107] Corresponding to the above-mentioned voice interaction method, this disclosure also proposes a voice interaction device. Figure 6 This is a structural schematic diagram of a voice interaction device 500 provided in an embodiment of this disclosure. Figure 4 As shown, it includes:

[0108] The recognition unit 510 is used to acquire the user's input voice command, recognize the voice command through an artificial intelligence model, and obtain the user's intent; the matching unit 520 is used to match the user's intent with the registered intents of multiple services to obtain the matching result, and determine the target service based on the matching result. The target service is used to execute at least one instruction corresponding to the user's intent; the registration unit 530 is used to generate a scene identifier corresponding to the user's intent under preset conditions and register the scene to obtain a registered scene. The registered scene includes: the intent type corresponding to the scene, the scene identifier, and the correspondence between the scene identifier and the target service; the preset conditions include intent feature data of the user intent being of a preset type, and the preset type includes at least one of dialogue interaction and data information acquisition; the distribution unit 540 is used to respond to the receipt of new preset type intent feature data, match the new preset type intent feature data with the registered scene, and if a corresponding target registered scene is matched, distribute the new preset type intent feature data to the target service corresponding to the target registered scene for execution according to the scene identifier of the target registered scene.

[0109] In some embodiments, the device further includes a prompting unit, which acquires a voice command input by the user, recognizes the voice command through an artificial intelligence model, and obtains the user intent. The arbitration module is used to: in response to recognizing multiple user intents from a single voice command, output first prompt information, the first prompt information being used to guide the user to select at least one user intent from the multiple user intents; acquire selection information input by the user, and determine the individual user intent of the voice command from the multiple user intents based on the selection information.

[0110] In some embodiments, before matching the user intent with the registration intents of multiple services to obtain a matching result, the registration module 530 is further configured to: obtain the intent corresponding to the service based on the functions and business scenarios of the multiple services; register at least one intent for each of the multiple services to obtain the registration intents of multiple services.

[0111] In some embodiments, the matching unit 520 is specifically configured to: match the registration intent of multiple services with the user intent; in response to the existence of two or more services whose registration intent matches the user intent, output a second prompt message, the second prompt message being used to guide the user to select at least one service from the two or more services; obtain the response information input by the user, determine the corresponding service based on the response intent of the response information, and determine the service as the target service.

[0112] In some embodiments, the registration scenario further includes scenario parameter information. The registration unit 530 is specifically used to: obtain the intent type under preset conditions and generate a scenario identifier corresponding to the user intent; establish a correspondence between the scenario identifier and the target service; and determine the scenario parameter information based on the user intent through the target service. The scenario parameter information includes one or more of the following: display device information corresponding to the scenario, identifier that needs to receive generative artificial intelligence data, participant whitelist, and registration time.

[0113] In some embodiments, the distribution unit 540 is specifically configured to: acquire new voice commands input by the user; recognize the new voice commands through an artificial intelligence big data model to obtain new preset type intention feature data of dialogue interaction; match the new preset type intention feature data of dialogue interaction with the intention types corresponding to multiple registration scenarios to obtain intention matching results; and determine the target registration scenario from multiple registration scenarios based on the intention matching results.

[0114] In some embodiments, the distribution unit 540 is further configured to: if the new preset type of intent feature data of dialogue interaction matches the intent type corresponding to two or more registration scenarios in multiple registration scenarios, obtain parameter information of the new preset type of intent feature data of dialogue interaction; match the parameter information of the new preset type of intent feature data of dialogue interaction with the scene parameter information of two or more registration scenarios to obtain parameter information matching results; and determine the target scenario from two or more registration scenarios based on the parameter information matching results.

[0115] In some embodiments, the distribution unit 540 is further configured to: respond to user intent feature data of a preset type and the preset type being data acquisition information, process the user intent and the scene identifier corresponding to the user intent through an artificial intelligence big data model to obtain generative artificial intelligence data, wherein the generative artificial intelligence data is bound to the scene identifier corresponding to the user intent; match the scene identifier corresponding to the user intent with the scene identifiers of multiple registered scenarios to obtain the identifier matching result; and determine the target registered scenario from the multiple registered scenarios based on the identifier matching result.

[0116] In some embodiments, the device further includes a deregistration unit, which is configured to: determine the display device corresponding to the registration scenario based on the scenario parameter information of the registration scenario; and deregister the registration scenario under a preset deregistration condition, wherein the preset deregistration condition includes at least one of the following: not receiving new preset type intent feature data matching the registration scenario for a preset period of time, or the user turning off the display device corresponding to the registration scenario.

[0117] In summary, according to the embodiments of this disclosure, the device, through an identification unit, a matching unit, a registration unit, and a distribution unit, distributes AI capabilities based on the scenario, thereby achieving full-vehicle, real-time, multi-user interaction of in-vehicle voice while supporting the generative content distribution of large models.

[0118] It should be noted that since the device embodiments of this disclosure correspond to the method embodiments described above, the foregoing explanations and descriptions of the method embodiments also apply to the device of this embodiment, and the principles are the same. For details not disclosed in the device embodiments, please refer to the method embodiments described above, and they will not be repeated in this disclosure.

[0119] According to embodiments of this disclosure, this disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0120] Figure 7 A schematic block diagram of an example electronic device 600 that can be used to implement embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0121] like Figure 7 As shown, device 600 includes a computing unit 601, which can perform various appropriate actions and processes based on a computer program stored in ROM (Read-Only Memory) 602 or a computer program loaded from storage unit 608 into RAM (Random Access Memory) 603. RAM 603 may also store various programs and data required for the operation of device 600. The computing unit 601, ROM 602, and RAM 603 are interconnected via bus 604. I / O (Input / Output) interface 605 is also connected to bus 604.

[0122] Multiple components in device 600 are connected to I / O interface 605, including: input unit 606, such as keyboard, mouse, etc.; output unit 607, such as various types of displays, speakers, etc.; storage unit 608, such as disk, optical disk, etc.; and voice interaction unit 609, such as network card, modem, wireless voice transceiver, etc. Voice interaction unit 609 allows device 600 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0123] The computing unit 601 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 601 include, but are not limited to, CPUs (Central Processing Units), GPUs (Graphics Processing Units), various special-purpose AI (Artificial Intelligence) computing chips, various computing units running machine learning model algorithms, DSPs (Digital Signal Processors), and any suitable processor, controller, microcontroller, etc. The computing unit 601 performs the various methods and processes described above, such as voice interaction methods. For example, in some embodiments, the voice interaction method may be implemented as a computer software program tangibly contained in a machine-readable medium, such as storage unit 608. In some embodiments, part or all of the computer program may be loaded and / or installed on device 600 via ROM 602 and / or voice interaction unit 609. When the computer program is loaded into RAM 603 and executed by the computing unit 601, one or more steps of the methods described above may be performed. Alternatively, in other embodiments, the computing unit 601 may be configured to perform the aforementioned voice interaction method by any other suitable means (e.g., by means of firmware).

[0124] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, FPGAs (Field Programmable Gate Arrays), ASICs (Application-Specific Integrated Circuits), ASSPs (Application-Specific Standard Products), SOCs (System-on-Chips), CPLDs (Complex Programmable Logic Devices), computer hardware, firmware, software, and / or combinations thereof. These various implementations may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0125] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0126] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, RAM, ROM, EPROM (Electrically Programmable Read-Only Memory) or flash memory, optical fiber, CD-ROM (Compact Disc Read-Only Memory), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0127] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (Cathode-Ray Tube) or LCD (Liquid Crystal Display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0128] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data voice interaction (e.g., voice interaction networks) of any form or medium. Examples of voice interaction networks include LANs (Local Area Networks), WANs (Wide Area Networks), the Internet, and blockchain networks.

[0129] Computer systems can include clients and servers. Clients and servers are generally geographically separated and typically interact via voice-interactive networks. The client-server relationship is established by computer programs running on the respective computers and having a client-server relationship with each other. Servers can be cloud servers, also known as cloud computing servers or cloud hosts, a hosting product within the cloud computing service ecosystem that addresses the management difficulties and weak business scalability inherent in traditional physical hosts and VPS (Virtual Private Server) services. Servers can also be servers for distributed systems or servers incorporating blockchain technology.

[0130] It's important to note that artificial intelligence (AI) is the study of enabling computers to simulate certain human thought processes and intelligent behaviors (such as learning, reasoning, thinking, and planning). It encompasses both hardware and software technologies. AI hardware technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, and big data processing. AI software technologies primarily include computer vision, speech recognition, natural language processing, machine learning / deep learning, big data processing, and knowledge graph technologies.

[0131] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this disclosure can be achieved, and this is not limited herein.

[0132] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A voice interaction method, characterized in that, The method includes: The system acquires user-input voice commands, and uses a large-scale artificial intelligence model to recognize the voice commands and obtain the user's intent. The user intent is matched with the registration intent of multiple services to obtain a matching result, and the target service is determined based on the matching result. The target service is used to execute at least one instruction corresponding to the user intent. Under preset conditions, a scene identifier corresponding to the user intent is generated and the scene is registered to obtain a registered scene. The registered scene includes: the intent type corresponding to the scene, the scene identifier, and the correspondence between the scene identifier and the target service. The preset conditions include intent feature data of the user intent being of a preset type, which includes at least one of dialogue interaction and data acquisition. The registered scene also includes scene parameter information, which includes one or more of the following: display device information corresponding to the scene, identifiers that need to receive generative artificial intelligence data, a whitelist of participants, and registration time. In response to receiving new intent feature data of the preset type, the new intent feature data of the preset type is matched with the registration scenario. If a corresponding target registration scenario is matched, the new intent feature data of the preset type is distributed to the target service execution corresponding to the target registration scenario according to the scenario identifier of the target registration scenario.

2. The method according to claim 1, characterized in that, After obtaining the user's voice input and recognizing the voice command using an artificial intelligence model to determine the user's intent, the method further includes: In response to recognizing multiple user intentions from a single voice command, a first prompt message is output, which guides the user to select at least one user intention from the multiple user intentions. Obtain user input selection information, and determine a single user intent for the voice command from the plurality of user intents based on the selection information.

3. The method according to claim 1, characterized in that, Before matching the user intent with the registration intents of multiple services to obtain a matching result, the method further includes: Based on the functions and business scenarios of multiple services, obtain the intent corresponding to the service; Register at least one intent for each of the multiple services to obtain the registration intents for the multiple services.

4. The method according to claim 3, characterized in that, The step of matching the user's intent with the registration intents of multiple services to obtain a matching result, and determining the target service based on the matching result, includes: Match the registration intent of the multiple services with the user intent; In response to the fact that the registration intent of two or more services among the plurality of services matches the user's intent, a second prompt message is output, which is used to guide the user to select at least one service from the two or more services. Obtain the user's input response information, determine the corresponding service based on the response intent of the response information, and identify the service as the target service.

5. The method according to claim 1, characterized in that, The step of generating a scene identifier corresponding to the user's intent under preset conditions and registering the scene to obtain the registered scene includes: Under preset conditions, the intent type is obtained, and a scene identifier corresponding to the user intent is generated; Establish the correspondence between the scene identifier and the target service; The target service determines the scene parameter information based on the user's intent.

6. The method according to claim 1, characterized in that, The step of responding to receiving new intent feature data of the preset type as dialogue interaction and matching it with the registration scenario based on the new intent feature data of the preset type includes: The system acquires new voice commands input by the user, identifies the new voice commands through the artificial intelligence big data model, and obtains new intent feature data of the preset type of dialogue interaction. The new intent feature data of the preset type of dialogue interaction is matched with the intent types corresponding to multiple registration scenarios to obtain intent matching results; Based on the intent matching results, the target registration scenario is determined from the multiple registration scenarios.

7. The method according to claim 6, characterized in that, The step of determining the target registration scenario from the multiple registration scenarios based on the intent matching result includes: If the new intent feature data of the preset type is dialogue interaction and matches the intent type corresponding to two or more registration scenarios in the multiple registration scenarios, obtain the parameter information of the new user intent; The parameter information of the new intent feature data of the preset type of dialogue interaction is matched with the scene parameter information of the two or more registered scenarios to obtain the parameter information matching result. Based on the matching results of the parameter information, the target scenario is determined from the two or more registered scenarios.

8. The method according to claim 1, characterized in that, The data information includes generative artificial intelligence data, and the process of matching the new intent feature data of the preset type with the registration scenario in response to receiving new intent feature data of the preset type includes: In response to the user intent being intent feature data of a preset type and the preset type being data acquisition information, the user intent and the scene identifier corresponding to the user intent are processed by the artificial intelligence big model to obtain generative artificial intelligence data, wherein the generative artificial intelligence data is bound to the scene identifier corresponding to the user intent; The scene identifier corresponding to the user intent is matched with the scene identifiers of multiple registered scenes to obtain the identifier matching result; Based on the identifier matching results, the target registration scenario is determined from the multiple registration scenarios.

9. The method according to claim 5, characterized in that, The method further includes: The display device corresponding to the registration scenario is determined based on the scenario parameter information of the registration scenario; Under preset cancellation conditions, the registration scenario is cancelled, wherein the preset cancellation conditions include at least one of the following: no new preset type of intent feature data matching the registration scenario is received for a preset period of time, or the user closes the display device corresponding to the registration scenario.

10. A voice interaction device, characterized in that, The device includes: The recognition unit is used to acquire the user's voice input, and to recognize the voice command through a large artificial intelligence model to obtain the user's intent; A matching unit is used to match the user intent with the registration intent of multiple services to obtain a matching result, and to determine a target service based on the matching result. The target service is used to execute at least one instruction corresponding to the user intent. A registration unit is used to generate a scene identifier corresponding to the user's intent under preset conditions and register the scene to obtain a registered scene. The registered scene includes: the intent type corresponding to the scene, the scene identifier, and the correspondence between the scene identifier and the target service. The preset conditions include intent feature data of the user intent being of a preset type, and the preset type includes at least one of dialogue interaction and data acquisition. The registered scene also includes scene parameter information, which includes one or more of the following: display device information corresponding to the scene, identifiers that need to receive generative artificial intelligence data, a whitelist of participants, and registration time. The distribution unit is configured to, in response to receiving new intent feature data of the preset type, match the new intent feature data of the preset type with the registration scenario, and if a corresponding target registration scenario is matched, distribute the intent feature data of the preset type to the target service execution corresponding to the target registration scenario according to the scenario identifier of the target registration scenario.

11. An electronic device, comprising: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-9.

12. A non-transitory computer-readable storage medium storing computer instructions, wherein, The computer instructions are used to cause the computer to perform the method according to any one of claims 1-9.

13. A computer program product comprising a computer program that, when executed by a processor, implements the method according to any one of claims 1-9.

Images