Vehicle-mounted voice assistant application module, equipment and vehicle

Abstract

The invention discloses a vehicle-mounted voice assistant application module, equipment and a vehicle. The vehicle-mounted voice assistant application module comprises a voice acquisition and recognition sub-module and a data output sub-module, the voice acquisition and recognition sub-module is used for acquiring user voice, and generating a voice request according to the user voice if the user voice comprises a wake-up word; sending the voice request to a corresponding request processing end, and sending response data fed back by the request processing end to the data output sub-module through a target protocol; wherein in the target protocol, the voice acquisition and recognition sub-module is used as a communication client, and the data output sub-module is used as a communication server; and the data output sub-module is used for receiving the response data sent by the voice acquisition and recognition sub-module and outputting the response data. Therefore, through module division, effective decoupling can be realized, the development cost and the maintenance cost are reduced, and the stability and the expandability of voice assistant application are also provided.

Description

Technical Field

[0001] The present invention relates to the field of data communication technology, and in particular to an in-vehicle voice assistant application module, equipment and vehicle. Background Art

[0002] At present, most major car entertainment applications are based on the open source Android system and Huawei's Hongmeng system. For Android system application development, the software development and design tend to design the car voice assistant application as a single application, and there is no clear division of modules within the application. The definition of basic capabilities and customized capabilities is not clear enough. There are disadvantages of low cohesion and high coupling in development, which is not conducive to unit testing, and its expansion capabilities are also limited. Summary of the invention

[0003] In view of the above problems, the present invention is proposed to provide an in-vehicle voice assistant application module, device and vehicle, which can achieve effective decoupling, reduce development costs and maintenance costs, and also provide stability and scalability of the in-vehicle voice assistant application.

[0004] According to a first aspect of the present invention, there is provided a vehicle-mounted voice assistant application module, comprising: a voice collection and recognition submodule and a data output submodule;

[0005] The voice collection and recognition submodule is used to collect user voice, and if the user voice includes a wake-up word, a voice request is generated according to the user voice; the voice request is sent to the corresponding request processing end, and the response data fed back by the request processing end is sent to the data output submodule through the target protocol; wherein, in the target protocol, the voice collection and recognition submodule acts as a communication client, and the data output submodule acts as a communication server;

[0006] The data output submodule is used to receive the response data sent by the voice collection and recognition submodule, and output the response data.

[0007] Optionally, the data output submodule includes a voice interface unit and a voice output unit;

[0008] The voice interface unit is used to call the response interface corresponding to the response data when the response data contains the information to be displayed, so as to display the information to be displayed through the response interface;

[0009] The voice output unit is used to play the audio information when the response data packet contains the audio information.

[0010] Optionally, the voice collection and recognition submodule is used to perform text recognition on the collected user voice to obtain semantic text; when a wake-up word is detected in the semantic text, the semantic text is parsed to obtain structured data, the structured data is encapsulated, a voice request is generated, and the voice request is sent to the corresponding request processing end.

[0011] Optionally, the voice collection and recognition submodule includes a recording unit and a voice recognition unit;

[0012] The recording unit is used to collect the user's voice and send the user's voice to the voice recognition unit;

[0013] The speech recognition unit is used to perform text recognition on the user speech to obtain the semantic text.

[0014] Optionally, the voice collection and recognition submodule includes a conversation service unit and a natural language understanding unit;

[0015] The conversation service unit is used to receive the semantic text, and when the wake-up word is detected in the semantic text, send the semantic text to the natural language understanding unit;

[0016] The natural language understanding unit is used to parse the semantic text to obtain structured data, encapsulate the structured data, construct a functional function for representing the user's voice intention, and generate a voice request based on the functional function.

[0017] Optionally, the voice request includes entity information, and the voice collection and recognition submodule further includes a plurality of agents;

[0018] The conversation service unit is further configured to determine, according to the entity information in the voice request, a target intelligent agent corresponding to the entity information among the multiple intelligent agents, and send the voice request to the target intelligent agent;

[0019] The target agent is used to send the voice request to the request processing end, and send the response data fed back by the request processing end to the data output submodule through the target protocol.

[0020] Optionally, the target protocol is the Android Binder protocol.

[0021] Optionally, the voice collection and recognition submodule is pre-bound to the data output submodule;

[0022] When the vehicle system is powered on and the voice collection and recognition submodule is started, the voice collection and recognition submodule triggers the data output submodule to start.

[0023] According to a second aspect of the present invention, there is provided an in-vehicle voice assistant device, characterized in that it includes the aforementioned in-vehicle voice assistant application module.

[0024] According to a third aspect of the present invention, a vehicle is provided, characterized in that it includes the aforementioned in-vehicle voice assistant device.

[0025] The above one or more technical solutions in the embodiments of this specification have at least the following technical effects:

[0026] The embodiments of this specification provide an in-vehicle voice assistant application module, device and vehicle, wherein the in-vehicle voice assistant application module includes a voice collection and recognition submodule and a data output submodule; the voice collection and recognition submodule is used to collect user voice, and if the user voice includes a wake-up word, a voice request is generated according to the user voice; the voice request is sent to the corresponding request processing end, and the response data fed back by the request processing end is sent to the data output submodule through the target protocol; wherein, in the target protocol, the voice collection and recognition submodule serves as a communication client, and the data output submodule serves as a communication server; the data output submodule is used to receive the response data sent by the voice collection and recognition submodule, and output the response data. In this way, through module division, the voice processing part and the display and playback part in the in-vehicle voice assistant application can be effectively decoupled, which reduces the development cost and maintenance cost, and also provides the stability and scalability of the voice assistant.

[0027] The above description is only an overview of the technical solution of the present invention. In order to more clearly understand the technical means of the present invention, it can be implemented according to the contents of the specification. In order to make the above and other purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present invention. Also, throughout the accompanying drawings, the same reference figures are used to represent the same components. In the drawings:

[0029] Figure 1 A communication schematic diagram of a vehicle-mounted voice assistant application module in an embodiment of the present invention is shown.

[0030] Figure 2 A block diagram of an in-vehicle voice assistant application module in an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0031] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0032] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0033] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, further definition and explanation thereof is not required in subsequent drawings.

[0034] In the description of the present invention, it is also necessary to explain that, unless otherwise clearly specified and limited, the terms "set", "install", "connect", and "connect" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two elements. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0035] At present, most major car entertainment applications are based on the open source Android system and Huawei's Hongmeng system. For Android system application development, the software development and design tend to design the car voice assistant application as a single application, and there is no clear division of modules within the application. The definition of basic capabilities and customized capabilities is not clear enough. There are disadvantages of low cohesion and high coupling in development, which is not conducive to unit testing, and its expansion capabilities are also limited.

[0036] Based on the above situation, the present invention combines the needs of voice with the actual development, comprehensively considers the actual requirements such as the basic capabilities of the application, customization capabilities, decoupling in the development process, unit testing, and expansion capabilities, and splits the in-vehicle voice assistant application module into multiple independent modules. Each module is responsible for specific basic capabilities and customization capabilities. And redefine the server and client in the communication in each module. In this way, effective decoupling can be achieved, which reduces development and maintenance costs, improves development efficiency, and also provides stability and scalability of the voice assistant.

[0037] Specifically, combined with Figure 1 , Figure 2As shown, an embodiment of the present invention provides a vehicle-mounted voice assistant application module, and the vehicle-mounted voice assistant application module includes: a voice collection and recognition submodule and a data output submodule.

[0038] In this embodiment, the voice collection and recognition submodule is used to collect user voice. If the user voice includes a wake-up word, a voice request is generated according to the user voice; the voice request is sent to the corresponding request processing end, and the response data fed back by the request processing end is sent to the data output submodule through the target protocol;

[0039] The data output submodule is used to receive the response data sent by the voice collection and recognition submodule, and output the response data.

[0040] The user voice refers to the voice uttered by the user in the external environment, for example, the voice uttered by the user when making a phone call, the voice uttered by the user when singing, the voice uttered by the user when speaking to the voice collection and recognition submodule, etc.

[0041] The wake-up word of this embodiment is pre-set and is used to wake up the voice collection and recognition submodule. The wake-up word can be "Hello, Xiaolan", "Xiaotu Xiaotu", etc. If it is detected that the user voice issued by the user includes the wake-up word, a voice request will be generated. A voice request refers to a request used to describe the user's voice intention. The voice request is sent to the request processing end, which can be a remote service or an application service on the car computer, for example, sending a voice request to the Moji Weather server. After receiving the voice request, the request processing end will process the voice request, generate response data, and then feed the response data back to the voice collection and recognition submodule. Response data refers to data generated by performing corresponding services based on voice requests. After that, the request processing end sends the response data to the data output submodule through the target protocol. The target protocol of this embodiment can be the Android Binder protocol.

[0042] The response data may include information to be displayed, such as weather data, stock data, etc. The data output submodule can display it by calling the corresponding response interface. The response data may also include audio information, and the data output submodule can output it by playing the audio information of the response data.

[0043] In this embodiment, the data output submodule includes a voice interface unit and a voice output unit.

[0044] The voice interface unit is used to call the response interface corresponding to the response data when the response data contains the information to be displayed, so as to display the information to be displayed through the response interface.

[0045] The voice output unit is used to play the audio information when the response data packet contains the audio information. In the voice output unit, a text-to-speech interface is provided internally and externally, which can be used to call the voice output unit for audio playback.

[0046] For example, the response data is news data, the voice interface unit calls the response interface corresponding to the news data to display the information to be displayed in the news data, and the voice output unit plays the audio information in the news data.

[0047] In this embodiment, the voice collection and recognition submodule may include a recording unit, a voice recognition unit, a conversation service unit, a natural language understanding unit, and multiple agents. The recording unit is connected to the voice recognition unit, the voice recognition unit is connected to the conversation service unit, and the conversation service unit is also connected to the natural language understanding unit and the multiple agents. It should be noted that the above-mentioned units can be set separately or partially integrated together.

[0048] In one embodiment, the voice collection and recognition submodule includes a recording unit and a voice recognition unit. The recording unit may be a recorder. The recording unit is used to collect user voice and send the user voice to the voice recognition unit; the voice recognition unit is used to recognize the user voice and obtain semantic text.

[0049] Among them, semantic text refers to text content generated based on user speech. After the user speaks, the recording unit will collect the user speech and send the collected user speech to the speech recognition unit for recognition to obtain semantic text. For example, if the user speaks "Today is a sunny day", the recording unit will collect the user speech and send the collected user speech to the speech recognition unit for recognition to obtain the semantic text of "Today is a sunny day".

[0050] In one embodiment, the voice collection and recognition submodule includes a conversation service unit and a natural language understanding unit. The conversation service unit is used to receive semantic text and detect whether the semantic text includes a wake-up word. If the wake-up word is detected in the semantic text, the voice enters the awake state from the dormant state, and the user voice in the awake state formally enters the active state. The subsequent conversation service unit can directly perform voice recognition and semantic text output.

[0051] The user voice of this embodiment is stateful. In a complete voice cycle, the state of the user voice can be divided into a dormant state, a wake-up state, a listening state, an input state, a recognition state, a semantic text output state, a voice structured output state, a play state, and the like. The dormant state refers to the state before the wake-up word is detected in the user voice, and the wake-up state refers to the state when the wake-up word is detected in the user voice. The listening state refers to the state of waiting to collect the user voice after the wake-up state. The input state refers to the state when the user voice is detected after the wake-up state. The recognition state refers to the state when the user voice is recognized after the input state. The semantic text output state refers to the state when the semantic text is output. The voice structured output state refers to the state when the semantic text is converted into structured data. The play state refers to the state when the voice output unit plays the voice. The voice interface unit can be used to describe the different states of the user voice.

[0052] In this embodiment, the voice interface unit includes a vehicle virtual assistant subunit, a voice recognition view subunit, a card subunit, and a sound wave subunit. The vehicle virtual assistant subunit is used for virtual assistant display, and the voice recognition view subunit is used to display the text content of the user's voice response in a typewriter-like streaming manner. The card subunit is used to display response data in a card-style response interface, such as weather data, stock data, etc.; it can also be used for interaction with users, and forwarding interaction requests to the voice collection and recognition submodule. The sound wave subunit is used to describe the direction and sound intensity of the voice input through sound waves. In the voice interface unit, the floating window is also extracted into a component, which provides an interface to the external body internally, and the interface provided to the external body can be used to serve external applications.

[0053] It should be noted that when a user voice is recognized, a complete semantic text is generated. However, this semantic text is not structured and cannot be recognized by the application. Therefore, the semantic text needs to be sent to the conversation service unit, and then the conversation service unit will send the semantic text to the natural language understanding (NLU) unit. The natural language understanding unit is used to parse the semantic text to obtain structured data, and then encapsulate the structured data to construct a function function for characterizing the user's voice intention, and then generate a voice request based on the function function.

[0054] In one embodiment, the voice collection and recognition submodule further includes a plurality of intelligent agents. The natural language understanding unit sends the voice request to the conversation service unit, and the conversation service unit sends the voice request to the corresponding intelligent agent.

[0055] Each agent describes different business functions, such as weather query, stock query, vehicle control, etc. The agent will forward the voice request to different request processing ends for processing.

[0056] In this embodiment, the request processing end can be a remote service or an on-board application service. For example, the weather query agent sends the voice request for weather query to the Moji Weather server; the vehicle control service agent sends the voice request for air conditioning control to the corresponding vehicle control service interface.

[0057] After receiving the response data fed back by the corresponding request processing end, the intelligent agent will send the response data to the voice interface unit and the voice output unit.

[0058] It should be noted that different voice requests will be sent to different agents for forwarding processing. In this embodiment, the voice request includes entity information, behavior information and target request parameters corresponding to the user voice.

[0059] The conversation service unit determines the target agent corresponding to the entity information among the multiple agents according to the entity information in the voice request, and sends the voice request to the target agent. It can be understood that a target relationship table is set in the conversation service unit, which includes entity information and agents corresponding to the entity information. Through the entity information of the voice request, the conversation service unit can match the agent corresponding to the voice request, and the corresponding agent is the target agent.

[0060] The target agent is used to send the voice request to the request processing end, and send the response data fed back by the request processing end to the voice interface unit and the voice output unit through the target protocol.

[0061] In order to facilitate understanding and implementation by those skilled in the art, this embodiment is described by the following examples:

[0062] Example 1

[0063] The semantic text is "Query the weather in Wuhan today", which is then parsed into structured data by the natural language understanding unit:

[0064] Entity information: weather;

[0065] Behavioral information: query;

[0066] Target request parameters: The geographical location is Wuhan, and the query time is today.

[0067] After the natural language understanding unit encapsulates this structured data into a voice request, it sends it to the conversation service unit. The conversation service unit determines the weather query agent from the agents as the target agent based on the entity information in the voice request. The conversation service unit sends the voice request to the target agent. The target agent sends the voice request to the Moji Weather server. The Moji Weather server generates response data based on the voice request, and then the Moji Weather server sends the generated response data to the target agent. The target agent obtains the response data about the weather query, which includes the information to be displayed and the audio information. The target agent sends the response data to the voice interface unit, and the voice interface unit calls up the response interface and displays it in the form of a card. At the same time, the target agent sends the response data to the voice output unit for broadcast.

[0068] Example 2

[0069] The semantic text is "Turn on the air conditioning", which is then parsed into structured data by the natural language understanding unit:

[0070] Entity information: air conditioning;

[0071] Behavior information: refrigeration switch;

[0072] Target Request Parameters: Open.

[0073] The natural language understanding unit encapsulates this structured data into a voice request and sends it to the conversation service unit. The conversation service unit determines the vehicle control service agent as the target agent from the agents based on the entity information in the voice request. The conversation service unit sends the voice request to the target agent. The target agent sends the voice request to the vehicle control service interface to operate the air conditioning switch. The vehicle control service interface feeds back the response data to the target agent. The target agent obtains the response data about the air conditioning control, and the target agent sends the response data to the voice interface unit and the voice output unit. The voice interface unit displays the corresponding text content of the response data in the form of a typewriter streaming on the screen, and the voice output unit broadcasts the response data.

[0074] It is worth mentioning that in the target protocol, the voice acquisition and recognition submodule acts as the client of communication, and the data output submodule acts as the server of communication; in this embodiment, the target protocol is the Android Binder protocol. The voice acquisition and recognition submodule is pre-bound to the data output submodule; when the vehicle system is powered on and the voice acquisition and recognition submodule is started, the voice acquisition and recognition submodule triggers the data output submodule to start.

[0075] It should be explained that in Binder communication, the first thing is to clarify the roles in the C / S architecture. That is, to clarify the client and the server. This embodiment proposes to use the voice collection and recognition submodule as the client in the communication, and the data output submodule as the server in the communication (that is, the voice interface unit and the voice output unit as the server in the communication), mainly based on the following considerations:

[0076] The voice collection work is completed by the voice collection and recognition submodule, and the application interaction requires the voice to be awakened by the wake-up word or the steering wheel control button. This requires that the process of the voice collection and recognition submodule should be started after the car system is started. After the voice wake-up, the virtual assistant and the voice-to-text view (Automatic Speech Recognition View) that guides the user will be presented, as well as the voice prompt broadcast. Based on this, it is more appropriate to start the data output submodule within the process of the voice collection and recognition submodule.

[0077] In terms of interface design, the data output submodule has no icon entry on the desktop, and there is no practical need for the module to start automatically after the system starts. Therefore, it is necessary to actively start the data output submodule process within the voice collection and recognition submodule process.

[0078] In the interaction process, the voice collection and recognition submodule first constructs a voice request and is the initiator of the communication. The voice collection and recognition submodule is more in line with the client role positioning of both parties in the communication.

[0079] During the interaction, the voice state switching maintained by the voice acquisition recognition submodule is very frequent, and the time dimension is tens of milliseconds. Similarly, in actual scenarios, voice request switching is also very fast. With such frequent state switching and request switching, if the voice acquisition recognition submodule is used as the server, the voice acquisition recognition submodule needs to send data to the data transmission submodule in the form of broadcast in a traditional communication mode, and the voice acquisition recognition submodule cannot receive the return information of the broadcast data, and the voice acquisition recognition submodule cannot know whether the data transmission submodule has received it or processed it. In addition, since there may be multiple clients, the server may need to lock or ensure thread synchronization when broadcasting, which may easily lead to untimely state switching of the voice interface submodule and even packet loss. Therefore, it is also technically necessary to define the voice acquisition recognition submodule as a client of communication, otherwise it is bound to cause performance and functional defects. If the voice acquisition recognition submodule is used as the client, the voice acquisition recognition submodule sends a request to the data output submodule. After the data output submodule receives and processes the request, it will also send a return message to the voice acquisition recognition submodule for confirmation.

[0080] Defining the data output submodule as the communication server can also meet the expansion requirements of the in-vehicle voice assistant application to provide integration capabilities from the technical architecture. For example, the voice interface unit exposes the existing floating window component to external applications; the voice output unit exposes the existing broadcast component to external applications.

[0081] For the customized capabilities of in-vehicle voice assistant applications, such as the scenario of voice-controlled media applications, the corresponding communication solution is the external media application as the server and the in-vehicle voice assistant application module as the client. During the startup phase, the in-vehicle voice assistant application binds to the server of the media application in a timely manner, starts the media process, and defines the data output sub-module as the server, which is also more in line with the overall software architecture design.

[0082] In summary, according to the division of the vehicle voice assistant application module and combined with the IPC mechanism of the Android system binder, the voice interface submodule and the conversion output submodule are defined as the server of the two communicating parties, and the voice collection and recognition submodule is defined as the client of the two communicating parties. When the vehicle is powered on and the Android system is started, the voice collection and recognition submodule process is started, and the voice interface submodule and the conversion output submodule services are timely bound inside the voice collection and recognition submodule process, and the voice interface submodule and the conversion output submodule processes are started.

[0083] In summary, an embodiment of this specification provides an in-vehicle voice assistant application module, which includes a voice collection and recognition submodule and a data output submodule; the voice collection and recognition submodule is used to collect user voice, and if the user voice includes a wake-up word, a voice request is generated according to the user voice; the voice request is sent to the corresponding request processing end, and the response data fed back by the request processing end is sent to the data output submodule through the target protocol; wherein, in the target protocol, the voice collection and recognition submodule serves as a communication client, and the data output submodule serves as a communication server; the data output submodule is used to receive the response data sent by the voice collection and recognition submodule, and output the response data. In this way, through module division, the voice processing part and the display and playback part in the in-vehicle voice assistant application can be effectively decoupled, which reduces the development cost and maintenance cost, and also provides the stability and scalability of the voice assistant.

[0084] Based on the same inventive concept, an embodiment of the present invention also provides an in-vehicle voice assistant device, including the aforementioned in-vehicle voice assistant application module.

[0085] Among them, the in-vehicle voice assistant application module includes: a voice collection and recognition sub-module and a data output sub-module.

[0086] The voice collection and recognition submodule is used to collect user voice. If the user voice includes a wake-up word, a voice request is generated according to the user voice; the voice request is sent to the corresponding request processing end, and the response data fed back by the request processing end is sent to the data output submodule through the target protocol;

[0087] The data output submodule is used to receive the response data sent by the voice collection and recognition submodule and output the response data. The response data can be played through audio and / or displayed through an interface.

[0088] Among them, the data output submodule includes a voice interface unit and a voice output unit.

[0089] The voice interface unit is used to call the response interface corresponding to the response data when the response data packet contains information to be displayed, so as to display the information to be displayed through the response interface; the voice output unit is used to play the audio information when the response data packet contains audio information.

[0090] For other contents in the in-vehicle voice assistant application module, please refer to the above content and will not be repeated here.

[0091] In summary, an in-vehicle voice assistant device provided in an embodiment of this specification includes an in-vehicle voice assistant application module, and the in-vehicle voice assistant application module includes a voice acquisition and recognition submodule and a data output submodule; the voice acquisition and recognition submodule is used to acquire user voice, and if the user voice includes a wake-up word, a voice request is generated according to the user voice; the voice request is sent to the corresponding request processing end, and the response data fed back by the request processing end is sent to the data output submodule through the target protocol; wherein, in the target protocol, the voice acquisition and recognition submodule serves as a client of communication, and the data output submodule serves as a server of communication; the data output submodule is used to receive the response data sent by the voice acquisition and recognition submodule, and output the response data. In this way, through module division, the voice processing part and the display and playback part in the in-vehicle voice assistant application can be effectively decoupled, which reduces the development cost and maintenance cost, and also provides the stability and scalability of the voice assistant.

[0092] Based on the same inventive concept, an embodiment of the present invention further provides a vehicle, comprising the aforementioned in-vehicle voice assistant device. The in-vehicle voice assistant device comprises an in-vehicle voice assistant application module.

[0093] Among them, the in-vehicle voice assistant application module includes: a voice collection and recognition sub-module and a data output sub-module.

[0094] The voice collection and recognition submodule is used to collect user voice. If the user voice includes a wake-up word, a voice request is generated according to the user voice; the voice request is sent to the corresponding request processing end, and the response data fed back by the request processing end is sent to the data output submodule through the target protocol;

[0095] The data output submodule is used to receive the response data sent by the voice collection and recognition submodule and output the response data. The response data can be played through audio and / or displayed through an interface.

[0096] Among them, the data output submodule includes a voice interface unit and a voice output unit.

[0097] The voice interface unit is used to call the response interface corresponding to the response data when the response data packet contains information to be displayed, so as to display the information to be displayed through the response interface; the voice output unit is used to play the audio information when the response data packet contains audio information.

[0098] In summary, an embodiment of the present specification provides a vehicle, in which an in-vehicle voice assistant application module includes a voice acquisition and recognition submodule and a data output submodule; the voice acquisition and recognition submodule is used to acquire user voice, and if the user voice includes a wake-up word, a voice request is generated according to the user voice; the voice request is sent to the corresponding request processing end, and the response data fed back by the request processing end is sent to the data output submodule through the target protocol; wherein, in the target protocol, the voice acquisition and recognition submodule serves as a client of communication, and the data output submodule serves as a server of communication; the data output submodule is used to receive the response data sent by the voice acquisition and recognition submodule, and output the response data. In this way, through module division, the voice processing part and the display and playback part in the in-vehicle voice assistant application can be effectively decoupled, which reduces the development cost and maintenance cost, and also provides the stability and scalability of the voice assistant.

[0099] The above are only various embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A vehicle-mounted voice assistant application module, characterized in that: include: Voice collection and recognition submodule and data output submodule; The voice collection and recognition submodule is used to collect user voice, and if the user voice includes a wake-up word, generate a voice request according to the user voice; Send the voice request to the corresponding request processing end, and send the response data fed back by the request processing end to the data output submodule through the target protocol; wherein, in the target protocol, the voice collection and recognition submodule acts as a communication client, and the data output submodule acts as a communication server; The data output submodule is used to receive the response data sent by the voice collection and recognition submodule, and output the response data.

2. The vehicle-mounted voice assistant application module according to claim 1, characterized in that: The data output submodule includes a voice interface unit and a voice output unit; The voice interface unit is used to call the response interface corresponding to the response data when the response data contains the information to be displayed, so as to display the information to be displayed through the response interface; The voice output unit is used to play the audio information when the response data packet contains the audio information.

3. The vehicle-mounted voice assistant application module according to claim 1, characterized in that: The voice collection and recognition submodule is used to perform text recognition on the collected user voice to obtain semantic text; when a wake-up word is detected in the semantic text, the semantic text is parsed to obtain structured data; The structured data is encapsulated to generate a voice request, and the voice request is sent to the corresponding request processing end.

4. The vehicle-mounted voice assistant application module according to claim 3, characterized in that: The voice collection and recognition submodule includes a recording unit and a voice recognition unit; The recording unit is used to collect the user's voice and send the user's voice to the voice recognition unit; The speech recognition unit is used to perform text recognition on the user speech to obtain the semantic text.

5. The vehicle-mounted voice assistant application module according to claim 3, characterized in that: The speech collection and recognition submodule includes a conversation service unit and a natural language understanding unit; The conversation service unit is used to receive the semantic text, and when the wake-up word is detected in the semantic text, send the semantic text to the natural language understanding unit; The natural language understanding unit is used to parse the semantic text to obtain structured data, encapsulate the structured data, construct a functional function for representing the user's voice intention, and generate a voice request based on the functional function.

6. The vehicle-mounted voice assistant application module according to claim 5, characterized in that: The voice request includes entity information, and the voice collection and recognition submodule also includes multiple agents; The conversation service unit is further configured to determine, according to the entity information in the voice request, a target intelligent agent corresponding to the entity information among the multiple intelligent agents, and send the voice request to the target intelligent agent; The target agent is used to send the voice request to the request processing end, and send the response data fed back by the request processing end to the data output submodule through the target protocol.

7. The vehicle-mounted voice assistant application module according to claim 1, characterized in that: The target protocol is the Android Binder protocol.

8. The vehicle-mounted voice assistant application module according to claim 1, characterized in that: The voice collection and recognition submodule is pre-bound to the data output submodule; When the vehicle system is powered on and the voice collection and recognition submodule is started, the voice collection and recognition submodule triggers the data output submodule to start.

9. A vehicle-mounted voice assistant device, characterized in that: Includes the in-vehicle voice assistant application module as described in any one of claims 1-8.

10. A vehicle, characterized in that: Including the in-vehicle voice assistant device as described in claim 9.

Images