Search

CN-122021639-A - Vehicle-mounted human-computer interaction method and device and electronic equipment

CN122021639ACN 122021639 ACN122021639 ACN 122021639ACN-122021639-A

Abstract

The application relates to the field of vehicles, and provides a vehicle-mounted man-machine interaction method, a device and electronic equipment, aiming at the problem of how to enable a vehicle-mounted terminal to accurately respond to the intention of a user and realize an application function expected by the user. The method comprises the steps of taking any one or more of vehicle data, driver data, environment and context data as a data source, obtaining a semantic text sequence according to the data source, wherein the semantic text sequence comprises one or more feature texts, the feature texts are used for describing any one or more of vehicle features, driver features, environment and context features, the feature texts are text character strings described by referring to natural language, the semantic text sequence is used as input of an intelligent model, obtaining a prediction result output by the intelligent model, and calling corresponding application services according to the prediction result. According to the method provided by the embodiment of the application, the prediction accuracy of the intelligent model can be improved, and the user experience of man-machine interaction can be improved.

Inventors

  • SHI ANWEI

Assignees

  • 联通智网科技股份有限公司

Dates

Publication Date
20260512
Application Date
20251224

Claims (10)

  1. 1. A vehicle-mounted man-machine interaction method, characterized in that the method is applied to electronic equipment, the method comprising: Taking any one or more of vehicle data, driver data, environment and context data as a data source, and acquiring a semantic text sequence according to the data source, wherein the semantic text sequence comprises one or more characteristic texts, the characteristic texts are used for describing any one or more of vehicle characteristics, driver characteristics, environment and context characteristics, and the characteristic texts are text character strings described by referring to natural language; taking the semantic text sequence as input of an intelligent model, and obtaining a prediction result output by the intelligent model; and calling the corresponding application service according to the prediction result.
  2. 2. The method of claim 1, wherein said obtaining a semantically text sequence from said data source comprises: When the numerical value of the data source changes beyond a preset threshold value and/or triggers a predefined event, corresponding feature text is generated and added to the semantical text sequence.
  3. 3. The method of claim 2, wherein the obtaining a semantically text sequence from the data source further comprises: And maintaining a context window, refreshing the context window according to a preset period, and placing the context window at the front end of the semantically text sequence, wherein the context window comprises one or more characteristic texts, and the characteristic texts contained in the context window are characteristic texts of a preset type.
  4. 4. The method according to claim 1, wherein the method further comprises: Deploying a pre-trained first intelligent model at a cloud end, training the first intelligent model, and generating an adapter file; and deploying a second intelligent model adapting to the vehicle hardware environment at the vehicle-mounted end, and loading the adapter file on the second intelligent model, wherein the second intelligent model is associated with the first intelligent model.
  5. 5. The method according to claim 4, wherein the method further comprises: and when the preset model adjustment conditions are met, performing personalized fine adjustment on the second intelligent model.
  6. 6. The method according to claim 4, wherein the method further comprises: Recording an interactive event chain; Desensitizing the interaction event chain to generate an optimization data packet, wherein the optimization data packet is used for optimizing the first intelligent model; and uploading the optimized data packet to a cloud end.
  7. 7. The method of any of claims 1-6, wherein the predicted outcome comprises a first predicted outcome for security and a second predicted outcome for user intent.
  8. 8. A vehicle-mounted human-machine interaction device, wherein the device is applied to an electronic device, the device comprising: The input acquisition module is used for acquiring a semantic text sequence by taking any one or more of vehicle data, driver data, environment data and context data as a data source according to the data source, wherein the semantic text sequence is a text character string described by referring to natural language; The model calling module is used for taking the semantic text sequence as the input of an intelligent model and obtaining a prediction result output by the intelligent model; and the service module is used for calling the corresponding application service according to the prediction result.
  9. 9. An electronic device comprising a memory and a processor; The processor is configured to execute instructions stored in the memory to cause the electronic device to perform the method of any one of claims 1-7.
  10. 10. A computer readable storage medium comprising computer program instructions which, when executed by a computer system, perform the method of any of claims 1-7.

Description

Vehicle-mounted human-computer interaction method and device and electronic equipment Technical Field The present application relates to the field of vehicles, and in particular, to a vehicle-mounted human-computer interaction method, device and electronic equipment. Background With the development of computer technology, more and more vehicles are equipped with a computer intelligent system (in-vehicle terminal). The user can perform man-machine interaction with the vehicle-mounted terminal so as to realize the vehicle-mounted application function. In a scenario where a user interacts with the vehicle-mounted terminal, the user needs to issue a detailed and accurate instruction to the vehicle-mounted terminal to enable the vehicle-mounted terminal to implement a vehicle-mounted application function (for example, adjusting a display interface of a vehicle-mounted display screen) expected by the user. However, in the driving scene, because the driver focuses on the driving of the vehicle, the driver cannot issue a detailed and accurate instruction to the vehicle-mounted terminal, which causes that the vehicle-mounted terminal cannot smoothly realize the application function expected by the user, thereby reducing the vehicle-mounted human-computer interaction experience. Therefore, a vehicle-mounted man-machine interaction method is needed, so that the vehicle-mounted terminal can accurately respond to the intention of a user, and the application function expected by the user is realized. Disclosure of Invention Aiming at the problem of how to enable the vehicle-mounted terminal to accurately respond to the intention of a user and realize the application function expected by the user, the application provides a vehicle-mounted human-computer interaction method, a device and electronic equipment, and also provides a computer program product and a computer readable storage medium. In the method of the embodiment of the application, based on the artificial intelligence (ARTIFICIAL INTELLIGENCE, AI) technology, the intelligent model is adopted to process the input data related to the application scene of the vehicle, so as to obtain the prediction result aiming at the intention of the vehicle user, and finally, the corresponding application function is realized according to the prediction result aiming at the vehicle user, thereby meeting the intention requirement of the vehicle user. Specifically, in the embodiment of the application, the intelligent model for realizing the prediction of the user intention is a unified artificial intelligent large model (Unified AI Large Model). According to the method of the embodiment of the application, a unified and universal AI large model is adopted as a central reasoning engine of the system to replace a traditional processing pipeline formed by a plurality of separated special models. To facilitate direct understanding and processing of the input data by the smart model, in one embodiment, the input data of the smart model is a text sequence with logical semantics (semantical text sequence (Semantic Text Sequence)). The semantical text sequence contains one or more feature texts, which are text strings described with reference to natural language. Specifically, the embodiment of the application adopts the following technical scheme: In a first aspect, the present application provides a vehicle-mounted human-computer interaction method, where the method is applied to an electronic device, and the method includes: Taking any one or more of vehicle data, driver data, environment and context data as a data source, and acquiring a semantic text sequence according to the data source, wherein the semantic text sequence comprises one or more characteristic texts, the characteristic texts are used for describing any one or more of vehicle characteristics, driver characteristics, environment and context characteristics, and the characteristic texts are text character strings described by referring to natural language; The semantic text sequence is used as input of an intelligent model, and a prediction result output by the intelligent model is obtained; And calling the corresponding application service according to the prediction result. According to the method of the first aspect, numerical modeling can be improved to the dimension of semantic understanding, prediction accuracy of the intelligent model is improved, and user experience of man-machine interaction is improved. According to the method of the first aspect, reasoning is performed based on the unified AI large model, the framework is simpler, global information association can be captured, and therefore the intelligent upper limit and expansibility are improved. According to the method, active prediction can be achieved, and compared with an equal passive system, the active prediction capability of the method enables interaction experience to be changed from 'people adapt to vehicles' to 'vehicle service people', and th