A method, device and vehicle for replying to in-vehicle voice interaction

Abstract

The present invention provides a method, device and vehicle for replying to in-vehicle voice interaction. The method identifies a first application corresponding to an application service voice instruction and calculates the correlation between the first application and other applications to obtain an associated application with the highest correlation with the first application. The method then combines a standard reply language corresponding to the first application and a personalized prompt language corresponding to the associated application to reply to the application service voice instruction. The method can broadcast more personalized related information to the user according to the user's voice instruction, thereby improving the user's experience and sense of interaction.

Description

Technical Field

[0001] The present invention relates to the technical field of in-vehicle entertainment audio technology, and in particular to a method, device and vehicle for replying to in-vehicle voice interaction. Background Art

[0002] With the widespread adoption of in-vehicle intelligent voice control, users can now use voice to query information on various entertainment, lifestyle, and vehicle services. They can also interact with users through differentiated voice commands through various methods, such as voice replication and algorithm optimization. However, existing solutions for voice response announcements (TTS) focus on making TTS more lifelike and punctuating sentences with greater emotion, focusing on the principle of "answering as many questions as needed." These solutions lack solutions that can provide personalized, combined, and effective responses in complex driving scenarios.

[0003] The current voice interaction solution is a one-question-one-answer solution. When users need multiple services, they need to initiate another inquiry to obtain them. The personalized broadcast is only a differentiated broadcast of the inquiry service and does not provide more information. Summary of the Invention

[0004] In view of this, the present invention provides a method, device and vehicle for replying to in-vehicle voice interaction, which can provide more relevant information recommendations to users in one inquiry, thereby improving the user's usage experience and sense of interaction.

[0005] A first aspect of the present invention provides a method for replying to an in-vehicle voice interaction, comprising:

[0006] Obtaining an application service voice command from a user, and identifying a first application corresponding to the application service voice command;

[0007] When the association recommendation broadcast service is enabled, the association analysis results between the first application and each second application in the user's historical driving behavior are obtained to obtain a positive correlation between the first application and each second application; wherein the second application is other system applications other than the first application;

[0008] selecting all second applications whose positive correlation with the first application is greater than a preset positive correlation threshold as candidate applications, and selecting the second application with the greatest positive correlation from all candidate applications that are not currently running as the associated application;

[0009] The application service voice instruction is replied by combining a preset standard reply language corresponding to the first application and a preset personalized prompt language corresponding to the associated application.

[0010] A second aspect of the present invention provides a vehicle-mounted voice interaction reply device, comprising:

[0011] a speech recognition module, configured to obtain a user's application service voice instruction and identify a first application corresponding to the application service voice instruction;

[0012] An application association analysis module, configured to obtain association analysis results between a first application and each second application in a user's historical driving behavior when the association recommendation broadcast service is enabled, and to determine a positive correlation between the first application and each second application; wherein the second application is a system application other than the first application;

[0013] an associated application determining module, configured to select all second applications whose positive correlation with the first application is greater than a preset positive correlation threshold as candidate applications, and select the second application with the greatest positive correlation from all candidate applications that are not currently running as the associated application;

[0014] The reply module is used to reply to the application service voice instruction by combining a preset standard reply language corresponding to the first application and a preset personalized prompt language corresponding to the associated application.

[0015] A third aspect of the present invention provides a vehicle, comprising a processor, a memory, and a computer program stored in the memory and capable of running on the processor, wherein the computer program, when executed by the processor, implements the above-mentioned in-vehicle voice interaction reply method.

[0016] Compared with the prior art, the in-vehicle voice interaction reply method, device, and vehicle provided by the present invention have the following beneficial effects:

[0017] The present invention identifies the first application corresponding to the application service voice instruction and calculates the correlation between the first application and other applications to obtain the associated application with the highest correlation with the first application, and then combines the standard reply language corresponding to the first application and the personalized prompt language corresponding to the associated application to reply to the application service voice instruction. It can broadcast more personalized related information to the user according to the user's voice instruction, thereby improving the user's experience and sense of interaction. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 It is a flow chart of an embodiment of the in-vehicle voice interaction reply method provided by the present invention. DETAILED DESCRIPTION

[0019] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0020] The present application provides a method and device for replying to in-vehicle voice interaction, wherein the vehicle terminal may be equipped with at least one microphone to support the pickup of sounds inside the vehicle.

[0021] In order to reduce the computing power of the vehicle computer system, the vehicle microphone can send the collected voice information to the cloud for semantic recognition. The cloud then performs application association analysis and sends the final calculation results (voice command reply) to the vehicle computer.

[0022] Reference below Figure 1 The reply method of the in-vehicle voice interaction provided by the embodiment of the present invention is described. The reply method of the in-vehicle voice interaction provided by the embodiment of the present invention is executed by the vehicle terminal, and includes steps S11 to S14:

[0023] S11, obtaining an application service voice instruction of a user, and identifying a first application corresponding to the application service voice instruction.

[0024] In this embodiment of the present invention, the vehicle's microphone picks up sounds inside the vehicle and, after picking up an application service voice command, parses the command, which carries key information corresponding to the application. For example, if the application service voice command is "Give me a seven-mile incense stick," the vehicle system recognizes the keyword "seven-mile incense stick" and associates it with the music application on the vehicle system, determining that the first application is a music application.

[0025] S12, when the association recommendation broadcast service is turned on, obtain the association analysis results of the user's first application and each second application in the historical driving behavior, and obtain the positive correlation between the first application and each second application; wherein the second application is other system applications other than the first application.

[0026] In an embodiment of the present invention, when a user turns on the association recommendation broadcast service, the vehicle system will obtain a positive correlation between the first application and each second application based on the association analysis result between the applications in the historical records.

[0027] Specifically, after each driving trip, the car computer will record the application services called by the user in all historical driving times, and perform correlation analysis between each application, such as the proportion of driving times in which application A and application B are run at the same time to all historical driving times, and the proportion of driving times in which application A and application B are run at the same time to the number of driving times in which application A is run, etc.

[0028] S13: All second applications whose positive correlation with the first application is greater than a preset positive correlation threshold are selected as candidate applications, and the second application with the greatest positive correlation is selected from all candidate applications that are not currently running as an associated application.

[0029] In an embodiment of the present invention, the greater the positive correlation, the higher the correlation between the second application and the first application. During the user's driving behavior, if the user chooses to call the service of the first application, the possibility of calling the second application at the same time is greater. In order to improve the activity of in-vehicle application use, the second application that is not running and has a greater positive correlation is selected as the associated application of the first application in this scenario.

[0030] Exemplarily, the positive correlation threshold is set to 1. In other embodiments, when the first application does not find a second application with a positive correlation greater than the positive correlation threshold, a corresponding hotspot prompt can be output. The hotspot prompt can be determined by the hotspot library. Specifically, the hotspot library broadcasts a sentence of the city or country based on the positioning, such as "Hefei Children's Hospital New District will start construction bidding", "The Ministry of Education responded to the problem of illustrations in the People's Education Edition of elementary school mathematics textbooks, and the People's Education Press apologized and made rectifications". Alternatively, if the first application does not find a second application with a positive correlation greater than the positive correlation threshold, no broadcast will be made, or a scene-prompt database will be established based on the current scene. For example, if the user voice outputs "Please turn on navigation", the car system recognizes the current time, the navigation route is the route home from get off work, and determines that the current scene belongs to the user going home, then establishes "Go home - Let's sing a cheerful song on the way home!" At this time, the car system replies "Let's sing a cheerful song on the way home".

[0031] S14, replying to the application service voice instruction by combining a preset standard reply language corresponding to the first application and a preset personalized prompt language corresponding to the associated application.

[0032] For example, after determining the associated application, a reply is made by combining the standard reply language of the first application and the personalized reply language of the associated application. For example, if the user inputs the voice command "Play me a song of Qilixiang" and it is determined that the associated application is navigation, the reply content is "Play Qilixiang for you. You can use navigation to get real-time traffic conditions on the way to work.

[0033] Specifically, the standard reply language for the first application can be obtained from a pre-established library of standard reply language for applications, which is used to indicate the corresponding relationship between applications and standard reply language. The personalized reply language for the associated application can be obtained from a pre-established library of personalized reply language for applications, which is used to indicate the relationship between application, user profile, and personalized prompt language. During implementation, the vehicle system determines the first user who issued the voice command based on the user profile of the first user, such as comprehensive information such as the user's personality, gender, and age. The user's personality can be obtained through machine learning by collecting information such as the user's speaking voice, tone of voice, and driving speed.

[0034] Based on the technical solution provided in the above embodiment, it identifies the first application corresponding to the application service voice instruction and calculates the correlation between the first application and other applications to obtain the associated application with the highest correlation with the first application, and then combines the standard reply language corresponding to the first application and the personalized prompt language corresponding to the associated application to reply to the application service voice instruction. It can broadcast more personalized related information to the user according to the user's voice instruction, thereby improving the user's experience and sense of interaction.

[0035] In an optional implementation, the positive correlation between the first application and each of the second applications is obtained by the following steps:

[0036] The ratio of the proportion of using the second application while using the first application in the historical driving times to the proportion of using the second application in the historical driving times is calculated to obtain a positive correlation between the first application and the second application.

[0037] Calculate the ratio of the proportion of using the second app while using the first app in the historical driving times to the proportion of using the second app in the historical driving times

[0038] That is, in the embodiment of the present invention, the lift is used to represent the positive correlation between applications.

[0039] Specifically, the car system applications will be continuously used by users during actual driving. Therefore, the user's support and confidence in the in-car applications can be calculated. Suppose there are applications A, B, C... in the car. The car system will calculate and update the support and confidence between each application after each trip.

[0040] Support is expressed as the number of times in which both apps A and B were used divided by the number of times apps were used in the driving history. P(A) represents the proportion of apps using app A. The support between apps A and B is Support = P(A&B). For example, if a driver drove 10 times and used both apps A and B 3 times, P(A&B) = 30%.

[0041] Confidence is expressed as the ratio of the number of times application A and application B are used simultaneously to the number of times application A is used. Confidence = P(A&B) / P(A)

[0042] When a user calls an application by voice, the cloud calculates the lift between the application and other applications, thereby obtaining a positive correlation between the application and other applications. That is, in the embodiment of the present invention, the positive correlation between applications is characterized by lift. For example, when driving, the user outputs a voice command: "Help me a Seven-Lixiang" to play a music application. At this time, the vehicle system calculates the lift between music application A and other applications. The lift represents the ratio of "the proportion of applications containing A that also contain application B" to "the proportion of applications containing application B". The formula is expressed as Lift = (P(A&B) / P(A)) / P(B) = P(A&B) / P(A) / P(B).

[0043] In an optional implementation, after obtaining the user's application service voice instruction, the method further includes:

[0044] Identifying the user voiceprint corresponding to the application service voice command and matching it with a pre-stored user voiceprint library to identify the first user who issued the application service voice command;

[0045] Then, the step of replying to the application service voice command by combining a preset standard reply language corresponding to the first application and a preset personalized prompt language corresponding to the associated application specifically includes:

[0046] Obtaining a first user standard reply language corresponding to the first application according to a preset application standard reply language library corresponding to the first user;

[0047] Obtaining a first user personalized prompt corresponding to the associated application according to a preset application personalized prompt library corresponding to the first user;

[0048] The first user standard reply language and the first user personalized prompt language are combined to reply to the application service voice instruction.

[0049] In the embodiment of the present invention, the user identity is also identified, and a specific personalized prompt is output based on the user identity, which can further improve the user's usage experience.

[0050] In an optional embodiment, the method further includes:

[0051] Selecting a personalized prompt corresponding to the associated application from a preset application personalized prompt library;

[0052] Furthermore, the personalized prompt library is used to indicate the corresponding relationship between the associated applications, user portraits and personalized prompts;

[0053] Alternatively, the personalized prompt library is used to indicate the corresponding relationship between associated applications, scenarios and personalized prompts.

[0054] In the embodiment of the present invention, the application personalized prompt library is obtained based on the application personalized prompt library, and the application personalized prompt library is pre-established and can be updated in real time.

[0055] In one example, if the application personalized prompt library shows the correspondence between associated applications, user portraits and personalized prompts, then after receiving the application service voice command, the vehicle system turns on the camera, calls the third-party age detection service, identifies the user's gender and age, and determines the user portrait. For example, when it is determined that the associated application is a navigation application, the application personalized prompt library is queried, and navigation and male are searched. The corresponding personalized prompt is "Although you are an experienced driver, it is recommended that you still turn on the navigation and listen."

[0056] In another example, if the application personalized prompt library shows the correspondence between associated applications, scenes and personalized prompts, then after obtaining the application service voice command, the vehicle system can determine the current scene based on weather, temperature and other information, query the application personalized prompt library, search for navigation and rain, and search for navigation and rain. The corresponding personalized prompt of the associated application is "The road is slippery in rainy days, it is recommended to use navigation to obtain real-time traffic conditions."

[0057] In an optional implementation, after identifying the first application corresponding to the application service voice instruction, the method further includes:

[0058] When the associated recommendation broadcast service is closed, a preset standard reply language corresponding to the first application is called to reply.

[0059] A second aspect of an embodiment of the present invention provides a reply device for in-vehicle voice interaction, including:

[0060] a speech recognition module, configured to obtain a user's application service voice instruction and identify a first application corresponding to the application service voice instruction;

[0061] An application association analysis module, configured to obtain association analysis results between a first application and each second application in a user's historical driving behavior when the association recommendation broadcast service is enabled, and to determine a positive correlation between the first application and each second application; wherein the second application is a system application other than the first application;

[0062] an associated application determining module, configured to select all second applications whose positive correlation with the first application is greater than a preset positive correlation threshold as candidate applications, and select the second application with the greatest positive correlation from all candidate applications that are not currently running as the associated application;

[0063] The reply module is used to reply to the application service voice instruction by combining a preset standard reply language corresponding to the first application and a preset personalized prompt language corresponding to the associated application.

[0064] In an optional implementation, the positive correlation between the first application and each of the second applications is obtained by the following steps:

[0065] The ratio of the proportion of using the second application while using the first application in the historical driving times to the proportion of using the second application in the historical driving times is calculated to obtain a positive correlation between the first application and the second application.

[0066] In an optional implementation, after obtaining the user's application service voice instruction, the method further includes:

[0067] Identifying the user voiceprint corresponding to the application service voice command and matching it with a pre-stored user voiceprint library to identify the first user who issued the application service voice command;

[0068] Then, the reply module is specifically used to:

[0069] Obtaining a first user standard reply language corresponding to the first application according to a preset application standard reply language library corresponding to the first user;

[0070] Obtaining a first user personalized prompt corresponding to the associated application according to a preset application personalized prompt library corresponding to the first user;

[0071] The first user standard reply language and the first user personalized prompt language are combined to reply to the application service voice instruction.

[0072] In an optional embodiment, the reply module is further configured to select a personalized prompt corresponding to the associated application from a preset application personalized prompt library;

[0073] Furthermore, the personalized prompt library is used to indicate the corresponding relationship between the associated applications, user portraits and personalized prompts;

[0074] Alternatively, the personalized prompt library is used to indicate the corresponding relationship between associated applications, scenarios and personalized prompts.

[0075] In an optional implementation, the reply module is further configured to call a preset standard reply corresponding to the first application to reply when the associated recommendation broadcast service is closed.

[0076] It should be noted that the in-vehicle voice interaction reply device provided in an embodiment of the present invention is used to execute all the steps and processes of the in-vehicle voice interaction reply method of the above embodiment. The working principles and beneficial effects of the two correspond one to one, and no further details will be given here.

[0077] In addition, the device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present embodiment.

[0078] A third aspect of an embodiment of the present invention provides a vehicle, comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, the reply control method for in-vehicle voice interaction provided in the above embodiment is implemented, for example Figure 1 S11~S13.

[0079] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A reply method for in-vehicle voice interaction, characterized in that: include: Obtaining an application service voice command from a user, and identifying a first application corresponding to the application service voice command; When the association recommendation broadcast service is enabled, the association analysis results between the first application and each second application in the user's historical driving behavior are obtained to obtain a positive correlation between the first application and each second application; wherein the second application is other system applications other than the first application; selecting all second applications whose positive correlation with the first application is greater than a preset positive correlation threshold as candidate applications, and selecting the second application with the greatest positive correlation from all candidate applications that are not currently running as the associated application; Replying to the application service voice command by combining a preset standard reply language corresponding to the first application and a preset personalized prompt language corresponding to the associated application; wherein, selecting a personalized prompt corresponding to the associated application from a preset application personalized prompt library; Furthermore, the personalized prompt library is used to indicate the corresponding relationship between the associated applications, user portraits and personalized prompts; Alternatively, the personalized prompt library is used to indicate the corresponding relationship between associated applications, scenarios and personalized prompts; After obtaining the user's application service voice instruction, the method further includes: Identifying the user voiceprint corresponding to the application service voice command and matching it with a pre-stored user voiceprint library to identify the first user who issued the application service voice command; Then, the step of replying to the application service voice command by combining a preset standard reply language corresponding to the first application and a preset personalized prompt language corresponding to the associated application specifically includes: Obtaining a first user standard reply language corresponding to the first application according to a preset application standard reply language library corresponding to the first user; Obtaining a first user personalized prompt corresponding to the associated application according to a preset application personalized prompt library corresponding to the first user; The first user standard reply language and the first user personalized prompt language are combined to reply to the application service voice instruction.

2. The vehicle-mounted voice interaction reply method according to claim 1, characterized in that: The positive correlation between the first application and each of the second applications is obtained by the following steps: The ratio of the proportion of using the second application while using the first application in the historical driving times to the proportion of using the second application in the historical driving times is calculated to obtain a positive correlation between the first application and the second application.

3. The vehicle-mounted voice interaction reply method according to claim 1, wherein: After identifying the first application corresponding to the application service voice instruction, the method further includes: When the associated recommendation broadcast service is closed, a preset standard reply language corresponding to the first application is called to reply.

4. A vehicle-mounted voice interaction reply device, characterized in that: include: a speech recognition module, configured to obtain a user's application service voice instruction and identify a first application corresponding to the application service voice instruction; An application association analysis module, configured to obtain association analysis results between a first application and each second application in a user's historical driving behavior when the association recommendation broadcast service is enabled, and to determine a positive correlation between the first application and each second application; wherein the second application is a system application other than the first application; an associated application determining module, configured to select all second applications whose positive correlation with the first application is greater than a preset positive correlation threshold as candidate applications, and select the second application with the greatest positive correlation from all candidate applications that are not currently running as the associated application; a reply module, configured to reply to the application service voice command by combining a preset standard reply language corresponding to the first application and a preset personalized prompt language corresponding to the associated application; After obtaining the user's application service voice instruction, the method further includes: Identifying the user voiceprint corresponding to the application service voice command and matching it with a pre-stored user voiceprint library to identify the first user who issued the application service voice command; Then, the reply module is specifically used to: Obtaining a first user standard reply language corresponding to the first application according to a preset application standard reply language library corresponding to the first user; Obtaining a first user personalized prompt corresponding to the associated application according to a preset application personalized prompt library corresponding to the first user; Replying to the application service voice command by combining the first user standard reply language and the first user personalized prompt language; The reply module is further configured to select a personalized prompt corresponding to the associated application from a preset application personalized prompt library; Furthermore, the personalized prompt library is used to indicate the relationship between the associated applications, user portraits and personalized prompts; Alternatively, the personalized prompt library is used to indicate the corresponding relationship between associated applications, scenarios and personalized prompts.

5. The vehicle-mounted voice interaction reply device according to claim 4, characterized in that: The positive correlation between the first application and each of the second applications is obtained by the following steps: The ratio of the proportion of using the second application while using the first application in the historical driving times to the proportion of using the second application in the historical driving times is calculated to obtain a positive correlation between the first application and the second application.

6. A vehicle, characterized in that: It includes a processor, a memory, and a computer program stored in the memory and capable of running on the processor. When the computer program is executed by the processor, it implements the in-vehicle voice interaction reply method as described in any one of claims 1 to 3.

Images