An interaction method of a virtual role and a display device

Abstract

The present disclosure provides a virtual character, a display device and an electronic device, and a method comprising: in response to a user inputting a first query message, determining a display subject of a display picture based on the first query message, the display subject comprising a target virtual character and / or a target browser; in the case where the display subject comprises the target virtual character, in response to the user inputting a second query message, determining voice content and / or appearance form of the target virtual character based on the second query message; in the case where the display subject comprises the target virtual character and the target browser, in response to the user inputting the second query message, determining a positional relationship between the target virtual character and the target browser based on the second query message; wherein the first query message and the second query message are different.

Description

Technical Field

[0001] This disclosure relates to the field of digital human technology, and more particularly to an interaction method, display device, and electronic device for a virtual character. Background Technology

[0002] With the rapid development of artificial intelligence and digital human technology, digital human interaction products are gradually being applied on a large scale in scenarios such as industrial operation and maintenance and information retrieval. Among these technologies, digital human interaction products fall into two incompatible categories: standalone intelligent assistants and web-embedded digital humans. Standalone intelligent assistants rely on built-in large models for dialogue interaction but cannot read webpage DOM data or perform webpage content retrieval. Web-based digital humans often connect to webpages via HTML tags or cross-process communication, requiring modification of the webpage source code for adaptation and integration. They can only be used with a single webpage and cannot function as independent assistants. Because these two types of digital human interaction products are only applicable to different scenarios, their applicability is limited. Summary of the Invention

[0003] In view of this, embodiments of the present disclosure provide an interaction method, display device, and electronic device for virtual characters, which solve the problems of seamless switching of interaction functions of existing data-human virtual characters, high device deployment and adaptation costs, and limited applicable scenarios.

[0004] The first aspect of this disclosure provides an interaction method for a virtual character, the method comprising: responding to a user inputting a first query message, determining a display subject of a display screen based on the first query message, the display subject including a target virtual character and / or a target browser; if the display subject includes a target virtual character, responding to a user inputting a second query message, determining the voice content and / or appearance of the target virtual character based on the second query message; if the display subject includes both the target virtual character and the target browser, responding to a user inputting a second query message, determining the positional relationship between the target virtual character and the target browser based on the second query message; wherein the first query message and the second query message are different.

[0005] According to embodiments of this disclosure, determining the display subject of the display screen based on the first query message includes: determining whether there is a web page content query instruction or a browser operation instruction in the first query message; if there is a web page content query instruction or a browser operation instruction in the first query message, driving the display screen to display the target virtual character and the target browser; if there is no web page content query instruction or browser operation instruction in the first query message, driving the display screen to display the target virtual character.

[0006] According to embodiments of this disclosure, when the display subject includes a target virtual character, in response to a user inputting a second query message, determining the voice content of the target virtual character based on the second query message includes: when the display subject only includes the target virtual character, inputting the second query message into a target large language model to obtain a first query result for the second query message; determining a target audio signal based on the first query result; and determining the voice content of the target virtual character based on the target audio signal.

[0007] According to embodiments of this disclosure, when the display subject includes a target virtual character, in response to a user inputting a second query message, determining the voice content of the target virtual character based on the second query message includes: when the display subject includes both the target virtual character and a target browser, inputting the second query message into a target large language model to generate target content for the target browser; determining a second query result for the second query message based on the target content of the target browser; wherein the second query result represents the extraction result of the target content in the target browser; determining a target audio signal based on the second query result; and determining the voice content of the target virtual character based on the target audio signal.

[0008] According to an embodiment of this disclosure, determining the appearance of a target virtual character based on a second query message includes: determining a target action signal of the target virtual character based on a target audio signal and a target virtual character template; wherein the target action signal characterizes the change in appearance of the target virtual character when playing audio content.

[0009] According to embodiments of this disclosure, driving a display screen to display a target virtual character includes: when the display subject includes the target virtual character, determining a first appearance configuration parameter of the target virtual character; driving the display screen to display the target virtual character based on the first appearance configuration parameter; when the display subject includes the target virtual character and a target browser, determining a second appearance configuration parameter of the target virtual character; and driving the display screen to display the target virtual character based on the second appearance configuration parameter; wherein the appearance configuration parameter includes at least one of the display position, size, and transparency of the target virtual character, and the first appearance configuration parameter and the second appearance configuration parameter are different.

[0010] According to embodiments of this disclosure, determining a second query result for a second query message based on the target content of the target browser includes: obtaining the initial content of the target browser before the second query message is input into the target large language model; and determining the second query result based on changes in the target content of the initial content.

[0011] According to embodiments of this disclosure, determining a second query result for a second query message based on the target content of the target browser includes: obtaining a first document object model structure corresponding to the Hypertext Markup Language of the target browser before the second query message is input into the target large language model; obtaining a second document object model structure corresponding to the Hypertext Markup Language of the target browser after the second query message is input into the target large language model; and determining a second query result for the second query message based on the first document object model structure and the second document object model structure.

[0012] According to embodiments of this disclosure, when the display subject includes a target virtual character and a target browser, in response to a user inputting a second query message, determining the positional relationship between the target virtual character and the target browser based on the second query message includes: determining the target content of the target browser based on the second query message; and determining the positional relationship between the target virtual character and the target browser based on the target content, wherein the positional relationship includes either a first positional relationship or a second positional relationship; wherein the first positional relationship indicates that the target virtual character and the target browser are superimposed on the display screen; and the second positional relationship indicates that the target virtual character and the target browser are separately displayed on the display screen.

[0013] According to embodiments of this disclosure, determining the positional relationship between a target virtual character and a target browser based on target content includes: determining whether the target browser is displayed in full screen when the target content is located in a first area of ​​the target browser window; if the target browser is displayed in full screen, determining that the target virtual character and the target browser are displayed in the display screen with a first positional relationship; if the target browser is not displayed in full screen, determining that the target virtual character and the target browser are displayed in the display screen with a second positional relationship; wherein, the first area is the area in the target browser window that is far from the target distance from the center of the window.

[0014] According to embodiments of this disclosure, determining the positional relationship between the target virtual character and the target browser based on the target content further includes: when the target content is located in a second area of ​​the target browser window, determining that the target virtual character and the target browser are displayed on the display screen in a first positional relationship; wherein, the second area is the area in the target browser window that is far from the center of the window.

[0015] According to embodiments of this disclosure, the method further includes: adjusting the appearance size and / or transparency of the target virtual character based on the target content, when the target virtual character and the target browser are displayed on the display screen in a first positional relationship.

[0016] A second aspect of this disclosure provides a display device, comprising: a processor configured to, in response to a user inputting a first query message, determine a display subject of a display screen based on the first query message, the display subject including a target virtual character and / or a target browser; if the display subject includes a target virtual character, in response to a user inputting a second query message, determine the voice content and / or appearance of the target virtual character based on the second query message; if the display subject includes both the target virtual character and the target browser, in response to a user inputting a second query message, determine the positional relationship between the target virtual character and the target browser based on the second query message; wherein the first query message and the second query message are different. A display configured to display the target virtual character based on the voice content and appearance of the target virtual character; the display also configured to display the target virtual character and the target browser based on the voice content and appearance of the target virtual character, and the positional relationship between the target virtual character and the target browser.

[0017] A third aspect of this disclosure provides an electronic device comprising: one or more processors; and a memory for storing one or more computer programs, wherein the one or more processors execute the one or more computer programs to implement the steps of any of the methods described above.

[0018] A fourth aspect of this disclosure is a computer-readable storage medium having a computer program or instructions stored thereon, which, when executed by a processor, implement the steps of any of the methods described above.

[0019] The virtual character interaction method, display device, and electronic device provided in the embodiments of this disclosure can achieve at least the following technical effects:

[0020] Based on different user query messages, the system adaptively determines whether to display a standalone virtual character or a virtual character paired with a browser, and accordingly adjusts the virtual character's voice content, appearance, and the positional relationship between the two. This enables flexible switching and integrated fusion of webpage-free intelligent interaction and webpage scene interaction modes, solving the problems of the separation between the independent intelligent assistant of the digital human and the embedded functions of the webpage, and the limited scene adaptation. At the same time, relying on the embedded browser to cooperate with the virtual character for interactive interaction, webpage content query and control can be realized without modifying the webpage source code, expanding the applicable scenarios of digital human interaction. Attached Figure Description

[0021] The above and other objects, features and advantages of this disclosure will become clearer from the following description of embodiments with reference to the accompanying drawings, in which:

[0022] Figure 1 The flowchart illustrating one of the interaction methods of a virtual character according to an embodiment of this application is shown schematically;

[0023] Figure 2 The flowchart illustrating a virtual character interaction method according to an embodiment of this application is shown in the diagram as follows:

[0024] Figure 3 The flowchart illustrating the third method of interaction of a virtual character according to an embodiment of this application is shown in the illustration.

[0025] Figure 4 The fourth flowchart illustrates an interaction method for a virtual character according to an embodiment of this application.

[0026] Figure 5 This schematic diagram illustrates a structural block diagram of a display device according to an embodiment of the present application;

[0027] Figure 6 A block diagram schematically illustrates an electronic device suitable for implementing an interaction method for a virtual character according to an embodiment of this application. Detailed Implementation

[0028] To make the objectives, technical solutions, and advantages of this disclosure clearer, the following detailed description is provided in conjunction with specific embodiments and accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this disclosure. All other embodiments obtained by those skilled in the art based on the embodiments of this disclosure without inventive effort are within the scope of protection of this disclosure.

[0029] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit this disclosure. The terms “comprising,” “including,” etc., as used herein indicate the presence of the stated features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0030] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein are to be interpreted in a manner consistent with the context of this specification, and not in an idealized or overly rigid way.

[0031] When using expressions such as "at least one of A, B and C", they should generally be interpreted in accordance with the meaning that is commonly understood by those skilled in the art (e.g., "a system having at least one of A, B and C" should include, but is not limited to, a system having A alone, a system having B alone, a system having C alone, a system having A and B, a system having A and C, a system having B and C, and / or a system having A, B and C, etc.).

[0032] The relevant technical terms used in the embodiments of this disclosure are explained below.

[0033] HyperText Markup Language (HTML) is a standard markup language for building web page structures. It is used to define various elements and layouts of web pages, such as text, images, buttons, and blocks, and is also the basic carrier of web page data in this embodiment of the disclosure.

[0034] The Document Object Model (DOM) is a node-tree structure formed by the browser after parsing an HTML document. All elements, text, and attributes within a webpage can be read and accessed through a program interface. In this embodiment of the disclosure, DOM data is obtained in read-only mode through the browser's native API to achieve webpage content parsing and information extraction.

[0035] An Application Programming Interface (API) is a predefined interface and set of rules for data exchange and function calls between software systems or between different modules within a software system.

[0036] The first aspect of this disclosure provides an interaction method for a virtual character, the method comprising: responding to a user inputting a first query message, determining a display subject of a display screen based on the first query message, the display subject including a target virtual character and / or a target browser; if the display subject includes the target virtual character, responding to a user inputting a second query message, determining the voice content and / or appearance of the target virtual character based on the second query message; if the display subject includes the target virtual character and the target browser, responding to a user inputting a second query message, determining the positional relationship between the target virtual character and the target browser based on the second query message; wherein the first query message and the second query message are different.

[0037] In the embodiments disclosed herein, the interaction between the virtual character and the user includes, but is not limited to, methods such as the user and the virtual character, or the browser, completing command input, information feedback, interface linkage, and content query.

[0038] In the embodiments of this disclosure, user input includes query commands and operation commands issued through voice, text, touch, or other means.

[0039] In the embodiments of this disclosure, query messages are various types of instruction information sent by the user, including ordinary dialogue messages, web page query instructions, browser operation instructions, etc., which are used to drive the execution of corresponding interactive logic.

[0040] By adopting the above method, the display subject is adaptively determined to be either a standalone virtual character or a virtual character paired with a browser based on different user query messages. The voice content, appearance, and positional relationship of the virtual character are adjusted accordingly. This achieves flexible switching and integrated fusion of webpage-free intelligent interaction and webpage scene interaction modes, solving the problems of the separation between the independent intelligent assistant of digital humans and the embedded functions of webpages, and the limited scene adaptation. At the same time, relying on the embedded browser to cooperate with the virtual character for interactive interaction, webpage content query and control can be realized without modifying the webpage source code, expanding the applicable scenarios of digital human interaction.

[0041] Figure 1 The illustration shows one of the flowcharts of an interaction method for a virtual character according to an embodiment of the present disclosure.

[0042] like Figure 1 As shown, the interaction methods for the virtual character include operations S110 to S130.

[0043] Operation S110 responds to the user inputting a first query message and determines the display subject of the display screen based on the first query message. The display subject includes the target virtual character and / or the target browser.

[0044] In the embodiments of this disclosure, the first query message is an instruction message initiated by the user for the first time to determine the interaction scenario. The first query message is used to distinguish whether the browser function needs to be enabled at present, so as to determine the main body of the interface display.

[0045] In the embodiments of this disclosure, the display screen is a visual display interface of the terminal device, used to present a virtual character, an embedded browser window, and a combination of the two.

[0046] In the embodiments of this disclosure, the display subject is a combination of functional carriers presented on the display screen, including two forms: displaying only the target virtual character and displaying both the target virtual character and the target browser.

[0047] In the embodiments of this disclosure, the target virtual character is a digital human that interacts with an embedded browser and is the subject that performs voice broadcasting, action display, and human-computer response.

[0048] In the embodiments of this disclosure, the virtual character, as a digital human image, can be a digital virtual image with the ability to display image, broadcast voice, and link facial expressions and actions. As a carrier for interaction with users, the virtual character can realize functions such as human-computer dialogue, web page content broadcasting, and interactive feedback.

[0049] In embodiments of this disclosure, the target browser may be an embedded browser that does not rely on the built-in webpage loading component of an external browser for rendering and displaying third-party webpages.

[0050] In operation S120, if the displayed subject includes the target virtual character, in response to the user inputting a second query message, the voice content and / or appearance of the target virtual character are determined based on the second query message.

[0051] In the embodiments of this disclosure, the second query message is a specific interaction instruction subsequently initiated by the user after the display subject is determined, which is used to trigger actual operations such as dialogue and Q&A, web page query, and interface layout adjustment.

[0052] In the embodiments of this disclosure, the voice content includes voice text broadcast by the virtual character, such as voice information such as question and answer replies, web page content interpretation, and key point summaries.

[0053] In the embodiments of this disclosure, the appearance is the visual representation of a virtual character on a display screen, including visual and dynamic effects such as display position, size, transparency, facial expressions, body movements, and lip movements.

[0054] In operation S130, when the displayed subject includes the target virtual character and the target browser, in response to the user inputting a second query message, the positional relationship between the target virtual character and the target browser is determined based on the second query message; wherein the first query message and the second query message are different.

[0055] In the embodiments of this disclosure, the positional relationship refers to the arrangement of the target virtual character and the target browser on the same display screen, which is divided into two modes: overlay display and separate display.

[0056] Specifically, the system responds to the first query message input by the user and determines the main subject displayed on the screen based on this message. The main subject is divided into two types: one containing only the target virtual character and the other containing both the target virtual character and the target browser. The two types of main subjects correspond to different subsequent interaction logics. When the main subject is only the target virtual character, the system continues to respond to the second query message input by the user and determines the voice content and appearance of the target virtual character based on this message. When the main subject contains both the target virtual character and the target browser, the system determines their positional relationship on the screen based on the second query message. The first query message and the second query message are different instructions, and the steps are sequentially related. The type of main subject determined in the previous step directly determines the direction of the subsequent interaction process, thereby realizing the simultaneous operation of two modes: pure interaction without a webpage and webpage-linked interaction.

[0057] For example, after the electronic device is powered on and enters the interactive interface, if the user enters the first query message "Enable webpage interaction function", the main display subject is determined to be a digital human and an embedded browser. Then, the user enters the second query message "Adjust digital human position", and the screen position relationship between the digital human and the browser is re-planned according to the instruction. If the user enters the first query message "Start dialogue", it is determined that the main display subject is only a digital human. Then, the user enters the second query message "Introduce relevant products of a certain company", and the corresponding broadcast voice is generated based on the query results of the second query message. At the same time, the digital human's facial expressions, body language and other appearance are adjusted to complete the interaction process.

[0058] According to an embodiment of this disclosure, in operation S110, the display subject of the display screen is determined based on the first query message, including operations S1110 to S1130.

[0059] Operation S1110 determines whether there is a webpage content query instruction or a browser operation instruction in the first query message.

[0060] In the embodiments of this disclosure, the webpage content query command is a command issued by the user to retrieve, read, and broadcast information inside a webpage, such as "read the page title" or "summarize the main content of the webpage".

[0061] In the embodiments of this disclosure, browser operation instructions are instructions issued by the user to control the running state of the embedded browser, such as page turning, window scaling, webpage redirection, page refresh, and other operation-type instructions.

[0062] In operation S1120, if a webpage content query instruction or browser operation instruction exists in the first query message, the driver displays the target virtual character and the target browser on the display screen.

[0063] In operation S1130, if there is no webpage content query instruction or browser operation instruction in the first query message, the driver displays the target virtual character on the display screen.

[0064] Specifically, after obtaining the first query message, the message content is first parsed to determine whether it contains web page content query instructions or browser operation instructions. This determination step is the basis for distinguishing between the two display subjects. If either of the above two types of instructions is identified, the target screen will simultaneously display the target virtual character and the target browser. If no relevant instructions are identified, only the target virtual character will be displayed on the screen, giving the switching of the display subject a clear judgment standard and ensuring the stability of automatic switching between the two modes.

[0065] For example, when a user enters the first query message "Help me interpret the technical parameters in this webpage," semantic parsing identifies that the content contains a webpage query instruction. According to preset rules, an embedded browser window is automatically loaded into the interface and displayed on the same screen as the digital human, entering the webpage interaction mode. When the user enters the first query message "Let's talk about the development trend of artificial intelligence," since no webpage or browser-related operation intent is recognized, the interface only retains the digital human image, the embedded browser remains in a dormant state, and it automatically enters a pure dialogue interaction mode, without requiring the user to manually click the mode switching button.

[0066] By employing the above method, the system determines user needs based on the presence of webpage-related instructions in the first query message, thereby driving the display of the appropriate subject. If webpage-related instructions are present, a virtual character and browser are displayed to meet the user's need for interaction with webpage information; otherwise, only the virtual character is displayed, focusing on natural language interaction with the virtual character to improve interaction efficiency and accurately meet the user's needs in different scenarios.

[0067] According to an embodiment of this disclosure, in operation S120, when the display subject includes a target virtual character, in response to the user inputting a second query message, the voice content of the target virtual character is determined based on the second query message, including operations S1210 to S1230.

[0068] In operation S1210, if the main display subject only includes the target virtual character, the second query message is input into the target large language model to obtain the first query result for the second query message.

[0069] In the embodiments of this disclosure, the target large language model is a pre-deployed general-purpose lightweight large model responsible for semantic understanding, dialogue response, web page content extraction, and intent parsing.

[0070] In embodiments of this disclosure, the first query result is the response text content generated by the large language model for ordinary user dialogue commands when only virtual characters are displayed.

[0071] Operation S1220 determines the target audio signal based on the first query result.

[0072] In embodiments of this disclosure, the target audio signal is a playable audio data stream generated by converting text response content, used to drive the virtual character to complete voice broadcasting.

[0073] Operation S1230 determines the voice content of the target virtual character based on the target audio signal.

[0074] Specifically, the second query message is input into the large language model as a specific question-and-answer instruction. The large language model completes semantic parsing and content creation and outputs the first query result. Then, the text is converted into an audio signal through speech synthesis technology to determine the broadcast voice content of the virtual character. Among them, the action signal of the virtual character is generated based on the target audio signal, so that the virtual character performs the corresponding posture and facial expression changes when playing the target audio signal.

[0075] For example, when the display screen is in the interactive mode that only shows the digital human, the user enters the second query message: "Briefly describe the role of the embedded browser in the digital human system". This question and answer instruction is sent to the built-in large language model. The model combines the knowledge base to generate the standardized text response content, which is the first query result. The speech synthesis module reads the text and encodes it into an audio signal. The digital human virtual character generates the corresponding posture and movement based on the audio signal, thereby completing the real-time voice broadcast.

[0076] By adopting the above method, when only the target virtual character is displayed, the second query message entered by the user is processed by the target large language model to generate corresponding voice content, thereby realizing the question-and-answer interaction between the user and the virtual character based on natural language.

[0077] According to an embodiment of this disclosure, in operation S120, when the display subject includes a target virtual character, in response to the user inputting a second query message, the voice content of the target virtual character is determined based on the second query message, including operations S1240 to S1270.

[0078] In operation S1240, when the display subject includes the target virtual character and the target browser, the second query message is input into the target large language model so that the target large language module generates the target content of the target browser.

[0079] In the embodiments of this disclosure, the target content of the target browser is the valid information in the webpage currently loaded by the embedded browser, including text, titles, paragraphs, data, and other webpage content that the user needs to query.

[0080] Operation S1250: Based on the target content of the target browser, determine the second query result for the second query message; wherein, the second query result represents the extraction result of the target content in the target browser.

[0081] In the embodiments of this disclosure, the second query result is the result obtained after extracting, organizing and summarizing the webpage content, and is the response content to the webpage query command.

[0082] Operation S1260 determines the target audio signal based on the second query result.

[0083] Operation S1270 determines the voice content of the target virtual character based on the target audio signal.

[0084] Specifically, when the display subject includes both the target virtual character and the target browser, the second query message is input into the target large language model through instruction parsing and webpage content localization. The large language model understands the user's query intent and locates the target content in the target browser that matches the instruction. Furthermore, the target content in the browser is extracted, organized, and summarized to generate the second query result. The second query result is not text generated out of thin air by the large model, but rather the result of extracting and processing the target content of the webpage, ensuring that the response content comes entirely from the real data of the webpage. Subsequently, the text content of the second query result is converted into a target audio signal, and the virtual character's broadcast voice is finally determined based on the target audio signal.

[0085] For example, the display screen simultaneously shows the target virtual character and the target browser, with the browser currently showing an e-commerce website. The user inputs a second query message, "What is the best-selling mobile phone on this website?" This query message is input into the target large-scale language model. The model generates target content based on the e-commerce website's webpage structure and content, i.e., filtering out the information on the best-selling mobile phones. Based on this target content, the second query result is determined to be key information such as the mobile phone's brand, model, and sales volume. A target audio signal is generated based on this result, with the signal content being "The best-selling mobile phone on this website is brand A, model B, with sales reaching XX units," thus determining the voice content of the target virtual character.

[0086] By employing the above method, when the displayed subject includes both the target virtual character and the target browser, the virtual character's voice content is generated by combining a large language model with the webpage content, achieving a deep integration of the virtual character and webpage information. Users can not only access webpage content but also better understand and utilize this content through the virtual character's voice explanations, solving the problems of independent intelligent assistants being unable to read webpage DOM data and the limited functionality of digital humans on webpages in related technologies.

[0087] According to an embodiment of this disclosure, in operation S120, the appearance of the target virtual character is determined based on the second query message, including operation S1280.

[0088] Operation S1280 determines the target action signal of the target virtual character based on the target audio signal and the target virtual character template; wherein, the target action signal represents the change in appearance of the target virtual character when playing voice content.

[0089] In the embodiments disclosed herein, the virtual character template is a preset digital human basic resource library, which includes basic materials such as standard expressions, lip-syncing animations, body movements, and image parameters.

[0090] In the embodiments of this disclosure, the target motion signal is a control command generated based on the audio content and speech rhythm, used to drive the virtual character to match lip movements, facial expressions and body movements to achieve audio-visual synchronization.

[0091] In the embodiments of this disclosure, the changes in appearance are dynamic actions performed synchronously by the virtual character during the speech delivery process, such as nodding, gestures, mouth opening and closing, and facial expression switching.

[0092] Specifically, regardless of whether the displayed subject only contains the target virtual character or includes both the target virtual character and the target browser, the system will generate corresponding query results based on the user's input second query message through the target large language model, and further determine the target audio signal based on these results. The target audio signal represents the voice information the virtual character wants to convey to the user, such as an answer to a question or a prompt for an operation. The target virtual character template is a pre-defined set of rules that defines the rules for the virtual character's appearance changes under different voice characteristics. These characteristics can include the emotional tone of the voice, the speed of the voice, and the type of content expressed by the voice. The acquired target audio signal is then analyzed in conjunction with the target virtual character template. For example, if the voice content of the target audio signal is a cheerful introduction to a tourist attraction, based on the settings for this type of voice characteristic in the target virtual character template, it is determined that the virtual character should exhibit relatively lively actions. By comprehensively analyzing the target audio signal and the target virtual character template, the target motion signal is determined. This motion signal precisely describes the specific changes in the virtual character's appearance while playing audio content, including the character's body movements, such as the direction and amplitude of arm movements, body rotation angles, and leg postures; changes in facial expressions, such as the degree of smiling or surprise; and possible special actions, such as nodding or shaking the head. For example, in the aforementioned tourist attraction introduction, the target motion signal might determine that the virtual character quickly points in the relevant direction while smiling brightly, leaning slightly forward, and gently shaking its head to complement the vivid audio description, making the virtual character's performance more lively and natural. Finally, based on the determined target motion signal, the target virtual character is driven to change its appearance according to these actions on the display screen, thus presenting an expressive virtual character image to the user and enhancing the user's immersion and enjoyment during interaction with the virtual character.

[0093] By adopting the above method, the appearance and movement changes of the virtual character are determined based on the target audio signal and the target virtual character template. This allows the virtual character to no longer be a static display when playing voice content, but to express the voice content more vividly through changes in body movements and facial expressions, thereby enhancing the immersion and fun of the user's interaction with the virtual character.

[0094] Figure 2 The second schematic diagram illustrates a flowchart of an interaction method for a virtual character according to an embodiment of the present disclosure.

[0095] According to an embodiment of this disclosure, driving the display screen to display the target virtual character includes operations S210 to S220.

[0096] In operation S210, if the main display subject includes the target virtual character, the first appearance configuration parameters of the target virtual character are determined; based on the first appearance configuration parameters, the display screen is driven to display the target virtual character.

[0097] In operation S220, when the display subjects include the target virtual character and the target browser, the second appearance configuration parameters of the target virtual character are determined; based on the second appearance configuration parameters, the display screen is driven to display the target virtual character.

[0098] The appearance configuration parameters include at least one of the target virtual character's display position, appearance size, and transparency, and the first appearance configuration parameter and the second appearance configuration parameter are different.

[0099] In the embodiments of this disclosure, the appearance configuration parameters are a set of configurable parameters that control the visual effects of the virtual character, including three core parameters: display position, appearance size, and transparency. Specifically, the first appearance configuration parameter is used when only the virtual character is displayed, corresponding to the conventional display form of the digital human; the second appearance configuration parameter is used when the virtual character and the browser are displayed simultaneously, corresponding to forms such as floating, semi-transparent, and adaptive positioning of the digital human.

[0100] Specifically, when the main display subject is only the target virtual character, a first appearance configuration parameter is determined. This parameter includes at least one of the following: the display position, size, and transparency of the target virtual character. For example, in a simple Q&A scenario, to allow users to focus on interacting with the virtual character, it is displayed in a larger, opaque position in the center of the screen to highlight it and facilitate user interaction. When the main display subject includes both the target virtual character and the target browser, different second appearance configuration parameters are determined to accommodate the simultaneous display of browser content. For example, in an e-commerce shopping scenario, the browser is used to display product details pages. To avoid obscuring too much product information, the virtual character might be displayed in a smaller size and with some transparency in a corner of the screen, allowing users to clearly view product information while still being able to interact with the virtual character at any time.

[0101] By employing the above method, different appearance configuration parameters are determined based on the different situations of the display subject to drive the display of virtual characters, enabling a reasonable layout of virtual characters and browser content in different scenarios. When only virtual characters are displayed, they are highlighted to facilitate interaction; when both virtual characters and the browser are displayed simultaneously, the display of browser content is taken into account, avoiding the virtual characters from obscuring too much useful information and improving the convenience for users to view information.

[0102] Figure 3 The third schematic diagram illustrates a flowchart of an interaction method for a virtual character according to an embodiment of the present disclosure.

[0103] According to an embodiment of this disclosure, in operation S1250, a second query result for the second query message is determined based on the target content of the target browser, including operations S310 to S320.

[0104] Operation S310 retrieves the initial content of the target browser before the second query message is input into the target large language model.

[0105] In the embodiments of this disclosure, the initial content is the complete content of the original webpage when the embedded browser has just finished loading and has not been modified.

[0106] Operation S320 determines the second query result based on the changes in the target content of the initial content.

[0107] In the embodiments of this disclosure, the changes in the target content refer to the dynamic changes in page text, nodes, and content, such as additions, modifications, and deletions, that occur after the webpage is loaded.

[0108] Specifically, before inputting the second query message into the target large language model, the initial content of the currently displayed page in the target browser is obtained. The second query message is then input into the target large language model, which analyzes and processes the target browser's content according to the query requirements, generating the target content. The changes between the initial content and the target content are compared to determine the newly added content in the target content, thus determining the second query result.

[0109] For example, the target browser is displaying a news website page. The initial content may include all page elements such as titles, summaries, images, and related links for various news articles. The user enters a second query, "Filter out today's sports news." The large language model filters the news website page content, extracting all relevant sections of today's sports news to obtain the target content. Entertainment and technology news sections from the initial content are removed from the target content, retaining only the titles and excerpts of today's sports news. The second query result is the specific content of today's filtered sports news, presented as a news list, including key information such as news titles, brief summaries, and publication times, satisfying the user's query needs.

[0110] By employing the above method to obtain the initial content of the target browser and determining the second query result based on its changes compared to the target content, information relevant to the user's query can be extracted from the webpage more accurately. The large language model processes webpage content by comparing changes, making the query results more targeted, avoiding interference from irrelevant information, and improving the accuracy of information extraction.

[0111] According to an embodiment of this disclosure, in operation S1250, determining the second query result for the second query message based on the target content of the target browser, further includes operations S330 to S350.

[0112] Operation S330 retrieves the first document object model structure corresponding to the Hypertext Markup Language of the target browser before the second query message is input into the target large language model.

[0113] In the embodiments of this disclosure, Hypertext Markup Language (HTML) serves as the standard markup language for building web pages and is the carrier of web page content and structure.

[0114] In the embodiments of this disclosure, the Document Object Model (DOM) structure is a node tree structure generated by the browser after parsing the webpage, used to describe the hierarchical relationship of all elements, text, and tags on the webpage; for example, the DOM structure can be obtained in a read-only manner to achieve webpage parsing in a non-intrusive manner.

[0115] In the embodiments of this disclosure, the first document object model structure is the original DOM structure of the browser before the user inputs a query command.

[0116] Specifically, before inputting the second query message into the target large language model, the first DOM structure corresponding to the current HTML page of the target browser is obtained. The DOM treats each element (such as tags, text, etc.) in the webpage as a node in a tree, displaying the structural hierarchy of the webpage. For example, for an e-commerce product page, the first DOM structure would contain the nodes corresponding to each element such as product images, prices, descriptions, and reviews, as well as their interrelationships.

[0117] Operation S340 retrieves the second document object model structure corresponding to the Hypertext Markup Language of the target browser after the target large language model is input in the second query message.

[0118] In the embodiments of this disclosure, the second document object model structure is the latest DOM structure updated in real time by the browser after the user inputs a query command.

[0119] Specifically, the second query message is input into the target large language model. After processing the webpage content, the model retrieves the second DOM structure corresponding to the HTML of the target browser page. For example, if the user inputs the second query message "extract the price and user reviews of this product", after processing by the large language model, the second DOM structure may only retain node information related to the product price and user reviews, while other nodes such as product image display and product parameter description may be removed or modified.

[0120] Operation S350 determines the second query result for the second query message based on the first document object model structure and the second document object model structure.

[0121] Specifically, by comparing the first and second DOM structures, changes such as the addition, deletion, and modification of nodes are analyzed. For example, the comparison might reveal that the second DOM structure has added highlighted attributes for price and review-related nodes, or that some nodes unrelated to price and reviews have been deleted. Based on the differences between the two DOM structures, a second query result is determined for the second query message. For example, the specific numerical value of the product price and the text content of user reviews are extracted from the second DOM structure to form the second query result, satisfying the user's query needs for product price and review information.

[0122] By using the above method, the first DOM structure and the second DOM structure corresponding to the target browser's HTML are obtained, and the second query result is determined based on the difference between the two. The DOM structure reflects the hierarchical relationship of web page elements, and by comparing its changes, the operation of the large language model on the web page content is determined, thereby more accurately determining the query result.

[0123] According to an embodiment of this disclosure, in operation S130, when the display subject includes a target virtual character and a target browser, in response to the user inputting a second query message, the positional relationship between the target virtual character and the target browser is determined based on the second query message, including operations S1310 to S1320.

[0124] Operation S1310 determines the target content of the target browser based on the second query message.

[0125] Operation S1320: Based on the target content, determine the positional relationship between the target virtual character and the target browser. The positional relationship includes either the first positional relationship or the second positional relationship.

[0126] The first positional relationship represents the target virtual character and the target browser being displayed superimposed on the display screen; the second positional relationship represents the target virtual character and the target browser being displayed separately on the display screen.

[0127] In the embodiments of this disclosure, when displayed in an overlay, the virtual character is overlaid on top of the browser window as a floating layer and does not leave the browser screen.

[0128] In the embodiments of this disclosure, when displayed separately, the virtual character and the browser window are displayed independently of each other and in separate areas, with no overlap between their images.

[0129] In embodiments of this disclosure, the browser window may be the visual window area occupied by an embedded browser in the display screen.

[0130] Specifically, based on the second query message entered by the user, the target content to be displayed by the target browser is determined. For example, in a map application scenario, if the user enters the second query message "show nearby restaurants," the target content is determined to be the map's label, name, distance, and other relevant information about nearby restaurants. When the target virtual character and the target browser are displayed on the screen, for example, in the map application scenario mentioned above, if the target content is a concentrated display of restaurants in a specific area on the map, to facilitate the user's intuitive view of restaurant locations and to hear the virtual character's introduction, the target virtual character and the target browser are presented in an overlay position. The virtual character may be overlaid in a corner of the map screen, and its transparency and size will be appropriately adjusted to avoid obscuring key map information. If the target content is more complex and requires the user to view all browser content, such as the map simultaneously displaying multiple information such as restaurants, hotels, and attractions, the target virtual character and the target browser are displayed separately on the screen. The virtual character appears next to the browser window, allowing the user to simultaneously view map information and related suggestions or explanations provided by the virtual character.

[0131] By adopting the above method, the target content of the target browser is determined based on the second query message, and the positional relationship between the virtual character and the target browser is determined accordingly. This allows for the selection of appropriate display layouts based on different content characteristics, such as overlay or separate display, making the display effect more in line with the user's viewing and interaction needs, and improving the user's visual experience when viewing the virtual character and web page content at the same time.

[0132] Figure 4 The fourth schematic diagram illustrates a flowchart of an interaction method for a virtual character according to an embodiment of the present disclosure.

[0133] According to an embodiment of this disclosure, in operation S1320, the positional relationship between the target virtual character and the target browser is determined based on the target content, including operations S410 to S430.

[0134] Operation S410 determines whether the target browser should be displayed in full screen when the target content is located in the first area of ​​the target browser window; where the first area is the area in the target browser window that is far from the center of the window.

[0135] If the target browser is displayed in full screen on the display screen, the target virtual character and the target browser are displayed on the display screen in the first position relationship.

[0136] If the target browser is not displayed in full screen on the display screen, the S430 operation determines that the target virtual character and the target browser are displayed on the display screen in a second positional relationship.

[0137] Specifically, when the target content is located at the edge of the browser window, away from the center (first area), if the browser is full-screen and there is no extra layout space, to avoid obscuring too much browser content and to ensure user interaction with the virtual character, the target virtual character and the browser are displayed in a first positional relationship, overlapping each other. If the user manually shrinks the browser to a windowed state and it is not displayed in full screen, leaving empty areas, to clearly display the browser content and the virtual character, the layout automatically switches to a second positional relationship, placing the digital human virtual image outside the browser window, displaying them separately. This layout is determined by combining both the content area and the window state, adapting to the interaction needs under different window states.

[0138] For example, on a video playback website, the target browser displays a video playback page. The user enters a second query message, "View related recommendations below the video," and the related recommendations are displayed in the first area. If the browser is playing the video in full screen, to avoid interfering with the user's viewing experience, the virtual character is displayed in a smaller size with a certain degree of transparency in the lower right corner of the video screen—this is the first positional relationship. If the browser is not in full screen, the virtual character is displayed on the right side of the browser—this is the second positional relationship, allowing the user to simultaneously view both the recommendations below the video and the information provided by the virtual character.

[0139] By employing the above method, when the target content is located in a specific area of ​​the target browser window, the positional relationship between the virtual character and the browser is determined based on whether the browser is in full-screen mode, thus achieving intelligent adjustment of the display layout. In full-screen mode, an overlay display is used to fully utilize screen space without obscuring too much content; in non-full-screen mode, a separate display is used to ensure clear content presentation.

[0140] According to an embodiment of this disclosure, in operation S1320, determining the positional relationship between the target virtual character and the target browser also includes operation S440.

[0141] Operation S440 determines that the target virtual character and the target browser are displayed on the screen in a first positional relationship when the target content is located in the second area of ​​the target browser window; wherein, the second area is the area in the target browser window that is far away from the center of the window.

[0142] Specifically, when the user's target content is concentrated in the edge area of ​​the browser window, the edge area is the second area away from the center of the window. Regardless of whether the current browser is displayed in full screen or in a windowed and minimized state, the target virtual character and the target browser are set to the primary position relationship and displayed on top of the webpage in an overlay form. Since the area where the target virtual character is located belongs to the secondary information area of ​​the browser window, the overlay form will not obscure important content in the browser.

[0143] By adopting the above method, when the target content is located in an area of ​​the target browser window that is far from the center of the window, the virtual character and the browser are superimposed, which enhances the interaction between the virtual character and the browser window without interfering with the user's normal browsing of the target content.

[0144] According to embodiments of this disclosure, the interaction method for virtual characters further includes: adjusting the appearance size and / or transparency of the target virtual character based on the target content, when the target virtual character and the target browser are displayed on the display screen in a first positional relationship.

[0145] In embodiments of this disclosure, adjusting the appearance size includes dynamically modifying the display size of the virtual character, supporting zooming in and zooming out.

[0146] In embodiments of this disclosure, adjusting transparency includes dynamically modifying the transparency of the virtual character's image to avoid obscuring webpage content when overlaid.

[0147] Specifically, when the target virtual character and the target browser are displayed overlaid on the screen, the size and / or transparency of the target virtual character are adjusted based on the target content. Once it is confirmed that the two are displayed overlaid, the target content in the target browser is analyzed, and the size and / or transparency of the virtual character are adjusted according to the importance of the target content and display requirements. This better demonstrates the relationship between the virtual character and the webpage content without affecting the user's viewing of the target browser content, thereby improving the overall user interaction experience.

[0148] For example, in a web search scenario, the target browser displays the search results page, with a virtual character overlaid on the page to provide search suggestions to the user. If the target content is the image display area in the search results, and the image details are crucial to the user, the transparency of the virtual character can be reduced to make it more transparent, or the size of the virtual character can be appropriately reduced, ensuring that the user can see both the search suggestions provided by the virtual character and clearly view the image details, so that the user can see both the search suggestions provided by the virtual character and clearly view the image details.

[0149] By adopting the above method, when the virtual character and the browser are displayed in an overlay manner, adjusting the appearance size and / or transparency of the virtual character based on the target content can better demonstrate the relationship between the virtual character and the web page content without affecting the user's viewing of the browser content. At the same time, by flexibly adjusting the appearance of the virtual character according to the importance of the target content and display requirements, the display effect is optimized and the user interaction experience is improved.

[0150] The following provides a detailed explanation of how to generate virtual character images.

[0151] In some scenarios, based on the acquired webpage DOM structure, the target content area corresponding to user interaction is located, and its characteristics are collected, including: area size (calculating the width, height, and proportion of the target content rectangle in the browser window); visual priority features (distinguishing target content into types such as core images and text, function buttons, form controls, and secondary annotation text, with core images and text and operation controls defined as high-priority content and pure annotation text defined as low-priority content); and area position features (confirming that the target content is located in the center, first edge, or second edge of the window). The current default rendering parameters of the virtual character are read, including baseline size, baseline transparency, and layer level. The virtual character layer is fixed at the upper layer of the browser, and webpage touch / click events are not intercepted. Simultaneously, the character's actions and lip-sync configurations are loaded.

[0152] Different adjustment levels are defined based on the priority, area proportion, and location of the target content. Adjustment methods include adjusting only transparency, adjusting only size, and adjusting both transparency and size simultaneously. All level parameters have preset threshold ranges: the virtual character size adjustment range is 50%~200% of the original size, and the transparency adjustment range is 10%~100%, ensuring interactive visibility while avoiding excessive occlusion. Specific judgment rules are as follows:

[0153] For high-priority target content, such as core images and text, function buttons, and forms; if the target content occupies 50% or more of the browser window: implement a strategy of simultaneously reducing the size and transparency to minimize occlusion; if the target content occupies 20% to 50% of the browser window (medium core area): prioritize reducing transparency, while keeping the size at the default; if the target content occupies less than 20% of the browser window (small area controls / text boxes): only fine-tune the transparency, while keeping the size unchanged.

[0154] For low-priority target content, such as notes, footers, and secondary side text, the default size and transparency of the virtual character are maintained regardless of the area size. The transparency is only slightly reduced when the character completely covers the area, and the size does not need to be modified.

[0155] In some scenarios, if users manually set the size and transparency of a virtual character via mouse dragging or voice commands, the system retains these user-defined parameters and only makes minor adjustments when these parameters would completely obscure the target content, prioritizing user actions. For example, if a user sets the character to 100% opaque, the system will automatically reduce the transparency to 70% when the character covers the core button, while maintaining the user-defined size.

[0156] In some scenarios, when the virtual idle time exceeds the preset time and automatically switches to the floating ball form, the parameters are adapted synchronously: the default size of the floating ball is 30% of the original character, and the transparency is fixed at 50%, regardless of the target content.

[0157] Based on the above-described virtual character interaction method, embodiments of this application also provide a display device. The following will be combined with... Figure 5 The display device is described in detail.

[0158] Figure 5 A schematic block diagram of a display device according to an embodiment of this application is shown.

[0159] like Figure 5 As shown, the display device 500 includes a processor 510 and a display 520.

[0160] The processor 510 is configured to, in response to a user inputting a first query message, determine the display subject of the display screen based on the first query message, the display subject including a target virtual character and / or a target browser; if the display subject includes a target virtual character, in response to a user inputting a second query message, determine the voice content and / or appearance of the target virtual character based on the second query message; if the display subject includes both a target virtual character and a target browser, in response to a user inputting a second query message, determine the positional relationship between the target virtual character and the target browser based on the second query message; wherein the first query message and the second query message are different.

[0161] Display 520 is used to display the target virtual character based on the voice content and appearance of the target virtual character; it is also used to display the target virtual character and the target browser based on the voice content and appearance of the target virtual character, as well as the positional relationship between the target virtual character and the target browser.

[0162] According to an embodiment of this application, the processor 510 is further configured to determine whether there is a webpage content query instruction or a browser operation instruction in the first query message; if there is a webpage content query instruction or a browser operation instruction in the first query message, drive the display screen to display the target virtual character and the target browser; if there is no webpage content query instruction or browser operation instruction in the first query message, drive the display screen to display the target virtual character.

[0163] According to an embodiment of this application, the processor 510 is further configured to, when the display subject only includes the target virtual character, input the second query message into the target large language model to obtain a first query result for the second query message; determine the target audio signal based on the first query result; and determine the voice content of the target virtual character based on the target audio signal.

[0164] According to an embodiment of this application, the processor 510 is further configured to, when the display subject includes a target virtual character and a target browser, input a second query message into a target large language model so that the target large language module generates target content for the target browser; determine a second query result for the second query message based on the target content of the target browser; wherein the second query result represents the extraction result of the target content in the target browser; determine a target audio signal based on the second query result; and determine the voice content of the target virtual character based on the target audio signal.

[0165] According to an embodiment of this application, the processor 510 is further configured to determine a target action signal of the target virtual character based on the target audio signal and the target virtual character template; wherein the target action signal characterizes the change in appearance of the target virtual character when playing voice content.

[0166] According to an embodiment of this application, the processor 510 is further configured to: determine a first appearance configuration parameter of the target virtual character when the display subject includes the target virtual character; drive the display screen to display the target virtual character based on the first appearance configuration parameter; determine a second appearance configuration parameter of the target virtual character when the display subject includes the target virtual character and the target browser; drive the display screen to display the target virtual character based on the second appearance configuration parameter; wherein the appearance configuration parameter includes at least one of the display position, appearance size and transparency of the target virtual character, and the first appearance configuration parameter and the second appearance configuration parameter are different.

[0167] According to an embodiment of this application, the processor 510 is further configured to obtain the initial content of the target browser before the second query message is input into the target large language model; and determine the second query result based on the changes in the target content of the initial content.

[0168] According to an embodiment of this application, the processor 510 is further configured to determine the target content of the target browser based on the second query message; and determine the positional relationship between the target virtual character and the target browser based on the target content, wherein the positional relationship includes any one of a first positional relationship and a second positional relationship; wherein the first positional relationship indicates that the target virtual character and the target browser are superimposed on the display screen; and the second positional relationship indicates that the target virtual character and the target browser are separately displayed on the display screen.

[0169] According to an embodiment of this application, the processor 510 is further configured to determine whether the target browser is displayed in full screen when the target content is located in the first area of ​​the target browser window; if the target browser is displayed in full screen, determine that the target virtual character and the target browser are displayed in the display screen in a first positional relationship; if the target browser is not displayed in full screen, determine that the target virtual character and the target browser are displayed in the display screen in a second positional relationship; wherein, the first area is the area in the target browser window that is far from the center target distance of the window.

[0170] According to an embodiment of this application, the processor 510 is further configured to determine that the target virtual character and the target browser are displayed on the display screen in a first positional relationship when the target content is located in the second area of ​​the target browser window; wherein, the second area is the area in the target browser window that is far away from the center of the window.

[0171] According to an embodiment of this application, the processor 510 is further configured to adjust the appearance size and / or transparency of the target virtual character based on the target content when the target virtual character and the target browser are displayed on the display screen in a first positional relationship.

[0172] Figure 6A block diagram schematically illustrates an electronic device suitable for implementing an interaction method for a virtual character according to an embodiment of this application.

[0173] like Figure 6 As shown, an electronic device 600 according to an embodiment of this application includes a processor 601, which can perform various appropriate actions and processes according to a program stored in a read-only memory 602 or a program loaded from a storage portion 608 into a random access memory 603. The processor 601 may include, for example, a general-purpose microprocessor, an instruction set processor and / or an associated chipset and / or a dedicated microprocessor. The processor 601 may also include onboard memory for caching purposes. The processor 601 may include a single processing unit or multiple processing units for executing different steps of the method flow according to an embodiment of this application.

[0174] Random access memory 603 stores various programs and data required for the operation of electronic device 600. Processor 601, read-only memory 602, and random access memory 603 are interconnected via bus 604. Processor 601 executes various steps of the method flow according to embodiments of this application by executing programs in read-only memory 602 and / or random access memory 603. It should be noted that the programs may also be stored in one or more memories other than read-only memory 602 and random access memory 603. Processor 601 may also execute various steps of the method flow according to embodiments of this application by executing programs stored in said one or more memories.

[0175] According to embodiments of this application, the electronic device 600 may further include an input / output interface 605, which is also connected to a bus 604. The electronic device 600 may also include one or more of the following components connected to the input / output interface 605: an input section 606 including a keyboard, mouse, etc.; an output section 607 including a cathode ray tube, liquid crystal display, etc., and a speaker, etc.; a storage section 608 including a hard disk, etc.; and a communication section 609 including a network interface card, such as a local area network card, modem, etc. The communication section 609 performs communication processing via a network such as the Internet. A drive 610 is also connected to the input / output interface 605 as needed. A removable medium 611, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on the drive 610 as needed so that computer programs read from it can be installed into the storage section 608 as needed.

[0176] Embodiments of this application also provide a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiments; or it may exist independently and not assembled into the device / apparatus / system. The computer-readable storage medium carries one or more programs, which, when executed, implement the method according to the embodiments of this application.

[0177] According to embodiments of this application, the computer-readable storage medium can be a non-volatile computer-readable storage medium, such as including but not limited to: portable computer disks, hard disks, random access memory, read-only memory, erasable programmable read-only memory, portable compact disk read-only memory, optical storage devices, magnetic storage devices, or any suitable combination thereof. In embodiments of this application, the computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. For example, according to embodiments of this application, the computer-readable storage medium may include the read-only memory 602 described above, and / or random access memory 603, and / or one or more memories other than read-only memory 602 and random access memory 603.

[0178] Embodiments of this application also include a computer program product comprising a computer program containing program code for performing the methods shown in the flowchart. When the computer program product is run on a computer system, the program code is used to cause the computer system to implement the methods provided in the embodiments of this application.

[0179] In one embodiment, the computer program may rely on a tangible storage medium such as an optical storage device or a magnetic storage device. In another embodiment, the computer program may also be transmitted and distributed in the form of signals over a network medium, and downloaded and installed via the communication section 609, and / or installed from the removable medium 611. The program code contained in the computer program can be transmitted using any suitable network medium, including but not limited to: wireless, wired, etc., or any suitable combination thereof.

[0180] In embodiments of this application, the computer program can be downloaded and installed from a network via communication section 609, and / or installed from removable medium 611. When the computer program is executed by processor 601, it performs the functions defined in the system of embodiments of this application. According to embodiments of this application, the systems, devices, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0181] According to embodiments of this application, program code for executing the computer programs provided in the embodiments of this application can be written in any combination of one or more programming languages. Specifically, these computational programs can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. The program code can be executed entirely on the user's computing device, partially on the user's device, partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0182] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0183] Those skilled in the art will understand that the features described in the various embodiments of this application can be combined and / or combined in various ways, even if such combinations or combinations are not explicitly described in this application. In particular, the features described in the various embodiments of this application can be combined and / or combined in various ways without departing from the spirit and teachings of this application. All such combinations and / or combinations fall within the scope of this application.

Claims

1. A method for interacting with a virtual character, characterized in that, The method includes: In response to a user inputting a first query message, the display subject of the displayed screen is determined based on the first query message, and the display subject includes the target virtual character and / or the target browser; When the display subject includes the target virtual character, in response to the user inputting a second query message, the voice content and / or appearance of the target virtual character are determined based on the second query message; When the display subject includes the target virtual character and the target browser, in response to the user inputting a second query message, the positional relationship between the target virtual character and the target browser is determined based on the second query message; The first query message and the second query message are different.

2. The method according to claim 1, characterized in that, The step of determining the display subject of the display screen based on the first query message includes: Determine whether the first query message contains a webpage content query instruction or a browser operation instruction; If the first query message contains the webpage content query instruction or the browser operation instruction, drive the display screen to display the target virtual character and the target browser; If the webpage content query instruction and the browser operation instruction are not present in the first query message, the display screen is driven to display the target virtual character.

3. The method according to claim 1, characterized in that, When the display subject includes the target virtual character, in response to user input of a second query message, determining the voice content of the target virtual character based on the second query message includes: When the main display subject only includes the target virtual character, the second query message is input into the target large language model to obtain the first query result for the second query message; Based on the first query result, the target audio signal is determined; Based on the target audio signal, the voice content of the target virtual character is determined.

4. The method according to claim 1, characterized in that, When the display subject includes the target virtual character, in response to user input of a second query message, determining the voice content of the target virtual character based on the second query message includes: When the display subject includes the target virtual character and the target browser, the second query message is input into the target large language model so that the target large language module generates the target content of the target browser; Based on the target content of the target browser, determine the second query result for the second query message; The second query result represents the extraction result of the target content in the target browser; Based on the second query result, the target audio signal is determined; Based on the target audio signal, the voice content of the target virtual character is determined.

5. The method according to claim 3 or 4, characterized in that, Based on the second query message, the appearance of the target virtual character is determined, including: Based on the target audio signal and the target virtual character template, the target action signal of the target virtual character is determined; The target motion signal represents the changes in the appearance of the target virtual character when playing voice content.

6. The method according to claim 2, characterized in that, The process of driving the display screen to show the target virtual character includes: When the display subject includes the target virtual character, a first appearance configuration parameter of the target virtual character is determined; based on the first appearance configuration parameter, the display screen is driven to display the target virtual character; When the display subject includes the target virtual character and the target browser, a second appearance configuration parameter of the target virtual character is determined; based on the second appearance configuration parameter, the display screen is driven to display the target virtual character; The appearance configuration parameters include at least one of the target virtual character's display position, size, and transparency, and the first appearance configuration parameter and the second appearance configuration parameter are different.

7. The method according to claim 4, characterized in that, Determining the second query result for the second query message based on the target content of the target browser includes: Before the second query message is input into the target large language model, the initial content of the target browser is obtained; The second query result is determined based on the changes in the target content of the initial content.

8. The method according to claim 4 or 7, characterized in that, Determining the second query result for the second query message based on the target content of the target browser includes: Before the second query message is input into the target large language model, the first document object model structure corresponding to the hypertext markup language of the target browser is obtained; After the target large language model is input in the second query message, the second document object model structure corresponding to the Hypertext Markup Language of the target browser is obtained; Based on the first document object model structure and the second document object model structure, a second query result is determined for the second query message.

9. The method according to claim 1, characterized in that, When the display subject includes the target virtual character and the target browser, in response to user input of a second query message, determining the positional relationship between the target virtual character and the target browser based on the second query message includes: Based on the second query message, the target content of the target browser is determined; Based on the target content, the positional relationship between the target virtual character and the target browser is determined, wherein the positional relationship includes any one of a first positional relationship and a second positional relationship; Wherein, the first positional relationship indicates that the target virtual character and the target browser are superimposed on the display screen; the second positional relationship indicates that the target virtual character and the target browser are displayed separately on the display screen.

10. The method according to claim 9, characterized in that, Determining the positional relationship between the target virtual character and the target browser based on the target content includes: If the target content is located in the first area of ​​the target browser window, determine whether the target browser is displayed in full screen on the display screen; If the target browser is displayed in full screen on the display screen, it is determined that the target virtual character and the target browser are displayed on the display screen in the first positional relationship; If the target browser is not displayed in full screen on the display screen, it is determined that the target virtual character and the target browser are displayed on the display screen in the second positional relationship; The first region is the area in the target browser window that is far from the center of the window target.

11. The method according to claim 9, characterized in that, The step of determining the positional relationship between the target virtual character and the target browser based on the target content further includes: When the target content is located in the second area of ​​the target browser window, the target virtual character and the target browser are determined to be displayed on the display screen in the first positional relationship; The second region is the area in the target browser window that is far from the center of the window target.

12. The method according to any one of claims 9 to 11, characterized in that, The method further includes: When the target virtual character and the target browser are displayed on the screen in the first positional relationship, the appearance size and / or transparency of the target virtual character are adjusted based on the target content.

13. A display device, characterized in that, The display device includes: A processor is configured to, in response to a user inputting a first query message, determine, based on the first query message, the display subject of a display screen, the display subject including a target virtual character and / or a target browser; if the display subject includes the target virtual character, in response to a user inputting a second query message, determine, based on the second query message, the voice content and / or appearance of the target virtual character; if the display subject includes both the target virtual character and the target browser, in response to a user inputting a second query message, determine, based on the second query message, the positional relationship between the target virtual character and the target browser; wherein the first query message and the second query message are different. A display for showing the target virtual character based on its voice content and appearance. The display is also used to display the target virtual character and the target browser based on the voice content and appearance of the target virtual character, as well as the positional relationship between the target virtual character and the target browser.

14. An electronic device, comprising: One or more processors; Memory, used to store one or more computer programs. The characteristic feature is that the one or more processors execute the one or more computer programs to implement the steps of the method according to any one of claims 1 to 12.

15. A computer-readable storage medium having a computer program or instructions stored thereon, characterized in that, When the computer program or instructions are executed by a processor, they implement the steps of the method according to any one of claims 1 to 12.

Images