Digital anchor interaction method and device, electronic equipment and storage medium

Abstract

This invention relates to the field of virtual digital technology and discloses a digital anchor interaction method, device, electronic device, and storage medium. The method includes: acquiring audience interaction information and converting the interaction information into interaction events; adjusting the digital anchor's state based on the interaction events, historical interaction events in a pre-stored contextual memory bank, and the digital anchor's historical state; generating and outputting the digital anchor's interaction content and actions based on the adjusted digital anchor's state; and updating the pre-stored contextual memory bank with the interaction events, the adjusted digital anchor's state, the interaction content and actions, and audience feedback on the interaction content and actions. This invention can comprehensively capture various forms of audience interaction, providing a rich data foundation for subsequent interaction event conversion, thereby more accurately understanding audience intentions and emotions.

Description

Technical Field

[0001] This invention relates to the field of virtual digital technology, specifically to digital anchor interaction methods, devices, electronic devices, and storage media. Background Technology

[0002] Digital anchors, as an emerging tool, are widely used in scenarios such as agricultural product marketing and agricultural knowledge popularization. However, existing digital anchor systems often exhibit mechanical content broadcasting and fixed response patterns when interacting with viewers, making it difficult to demonstrate human emotional expression and deep psychological states like real anchors. This limits their ability to establish emotional connections with viewers and enhance fan loyalty.

[0003] Specifically, against the backdrop of the rapid development of live-streaming models for "new farmers" (referring to new types of rural talents who, compared to traditional farmers, utilize modern scientific and technological knowledge, new concepts, and new tools to engage in agricultural production, operation, service, or management in areas such as rural development, construction, and governance), digital anchors are widely used in scenarios such as agricultural product marketing and agricultural knowledge popularization. They can achieve all-day content broadcasting and basic interaction with an anthropomorphic image. However, current digital anchor systems mostly operate based on preset scripts and simple response logic. Although they have basic voice and image interaction functions, they lack humanized emotional expression and deep psychological state presentation during live-streaming interactions. They are unable to convey a personalized, emotional, and immersive communication experience through language, expressions, and interactive behaviors like real anchors, thus limiting their ability to establish emotional connections with viewers and enhance fan loyalty and trust.

[0004] Existing digital broadcasting systems primarily exhibit mechanical content delivery and fixed responses during interactions, lacking intrinsic and continuous psychological state modeling and emotional evolution mechanisms. Their interactive behavior fails to generate coherent, anthropomorphic psychological activities in real time based on the communication context and audience feedback, and to naturally externalize these activities into verbal, facial, or behavioral responses. This results in emotionally hollow and stiff interactions during live streams. Especially in highly interactive scenarios such as live chat and real-time comments, the system struggles to simulate the immediate psychological reactions and emotional expressions of human broadcasters, failing to proactively imbue the live stream with personalized thoughts, emotions, and approachability. This limits the depth of interaction, audience engagement, and dissemination effectiveness of live streams, failing to meet the urgent needs of new business models for intelligent and emotionally resonant digital broadcasters. Summary of the Invention

[0005] This invention provides a digital anchor interaction method, device, electronic device, and storage medium to solve the problems of existing digital anchor interaction being mechanized and lacking emotional expression.

[0006] In a first aspect, the present invention provides a digital anchor interaction method, the method comprising:

[0007] Acquire audience interaction information and transform it into interactive events; Adjust the digital anchor's state based on interactive events, historical interactive events in the pre-stored context memory bank, and the digital anchor's historical state; Based on the adjusted status of the digital anchor, generate and output the interactive content and actions of the digital anchor; Update the pre-stored contextual memory bank with interactive events, the adjusted status of the digital anchor, interactive content and actions, and audience feedback on the interactive content and actions.

[0008] This invention provides a digital anchor interaction method that, by constructing a continuously evolving interaction mechanism, enables digital anchors to dynamically adjust their own state based on real-time interaction content and historical communication experience. On this basis, it generates multimodal responses that are consistent in semantics and behavior, and incorporates complete interaction process data into the memory system to optimize subsequent performance. This significantly improves the interaction coherence, context adaptability, and anthropomorphism of digital anchors in continuous dialogues, enabling more human and emotional interaction between digital anchors and viewers. It overcomes the limitations of existing digital anchors' mechanical and fixed responses, effectively enhancing the depth of interaction and viewer engagement.

[0009] In one alternative implementation, acquiring audience interaction information includes: Collect and identify viewers' voice messages, text comments, and virtual gift information as viewer interaction data; virtual gift information includes the value and type of the virtual gift.

[0010] The above technical solutions can comprehensively capture various forms of audience interaction, providing a rich data foundation for subsequent interactive event transformation, thereby enabling a more accurate understanding of audience intentions and emotions.

[0011] In one alternative implementation, converting interactive information into interactive events includes: The system converts live chat audio into audio-text, performs semantic analysis and intent recognition on the audio-text and text comments, and extracts audience comment keywords and potential intents. Based on the value and type of the virtual gifts given, an emotion recognizer is used to identify the audience's emotions. The audience comments keywords, underlying intentions, and audience emotions are integrated and transformed into interactive events.

[0012] The above technical solution can deeply process and integrate raw, multimodal audience interaction information to form structured interactive events, providing accurate input for digital anchors to adjust their state and ensuring the accuracy and relevance of interactive responses.

[0013] In one optional implementation, the historical interaction events include: historical interaction events within a preset time period stored in a time sequence and the corresponding historical behaviors of the digital anchor; the historical state of the digital anchor includes: the emotion value used to quantify the emotional dimension of the digital anchor within the preset time period and the focus of the digital anchor. Based on interactive events, historical interactive events in the pre-stored contextual memory bank, and the digital anchor's historical state, adjust the digital anchor's state, including: Based on interactive events, historical interactive events, the digital anchor's historical behavior, emotional value, and focus, the updated state parameters of the digital anchor are calculated through preset state transition rules or strategy models, and the digital anchor's current state is adjusted to the state corresponding to the updated state parameters in a smooth transition manner.

[0014] In the above technical solution, historical interaction events can provide historical context references for the digital anchor's state adjustment, making its interactive behavior more coherent and context-relevant, avoiding isolated and memoryless responses, and providing more refined internal state parameters for the digital anchor's state adjustment, enabling it to simulate the emotional changes and attention shifts of human anchors, thereby achieving more natural and personalized interactions.

[0015] In one optional implementation, based on the adjusted state of the digital anchor, interactive content and actions of the digital anchor are generated and output, including: Generate language response text that matches the adjusted digital anchor's status, and generate facial expression animation and lip-sync animation corresponding to the language response text; Select the appropriate body movement from the preset body movement library; The system performs time-coordinated scheduling and synchronous output of verbal response text, facial expression animation, lip-sync animation, and body movements. This technical solution generates verbal response text matching the adjusted state, along with corresponding facial expressions and lip-sync animations. Combined with body movements selected from a preset library, and through time-coordinated scheduling, it achieves synchronous output of multimodal elements, effectively ensuring the consistency of content, the coordination of expression, and the naturalness of behavior in the digital anchor's interactive performance.

[0016] In one optional implementation, the verbal response text, facial expression animation, lip-sync animation, and body movements are time-coordinated and output synchronously, including: Obtain the display area information of the current live broadcast screen, and adjust the body movements according to the display area information so that the body movements are fully presented within the display area; Perform time synchronization and emotional semantic consistency checks on language response text, facial expression animation, lip-sync animation, and body movements; The digital anchor model synchronously outputs adjusted and consistent language response text, facial expression animation, lip-sync animation, and body movements.

[0017] In the above technical solution, the body movements are adaptively adjusted according to the display area information to ensure their complete presentation within the screen. The language, facial expressions and movements are checked for time synchronization and emotional semantic consistency. Ultimately, the model is driven to achieve coordinated output of multimodal elements, which significantly improves the naturalness and fluency of digital anchor interaction, screen adaptability and accuracy of emotional expression.

[0018] In one alternative implementation, audience feedback on interactive content and actions includes: audience comments and emotional inclinations, audience likes, the number of gifts given, and the duration of audience engagement.

[0019] The above technical solution collects audience comments and their emotional tendencies, the number of likes and gifts, and the duration of their interaction as feedback information, thereby achieving a multi-dimensional quantitative evaluation of audience interaction responses. This provides a comprehensive and objective basis for updating the contextual memory bank, and supports the continuous optimization and precise adjustment of digital anchor interaction strategies.

[0020] Secondly, the present invention provides a digital broadcaster interactive device, the device comprising: The information acquisition and conversion module is used to acquire audience interaction information and convert it into interactive events; The status adjustment module is used to adjust the status of the digital anchor based on interactive events, historical interactive events in the pre-stored context memory bank, and the digital anchor's historical status. The interactive content and action generation module is used to generate and output the interactive content and actions of the digital anchor based on the adjusted state of the digital anchor. The context memory bank update module is used to update the pre-stored context memory bank with interactive events, the adjusted status of the digital anchor, interactive content and actions, and audience feedback on the interactive content and actions.

[0021] Thirdly, the present invention provides an electronic device, comprising: a memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the computer instructions to perform the digital anchor interaction method described in the first aspect or any corresponding embodiment thereof.

[0022] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions for causing a computer to execute the digital anchor interaction method described in the first aspect or any corresponding embodiment thereof.

[0023] Fifthly, the present invention provides a computer program product, including computer instructions for causing a computer to execute the digital anchor interaction method described in the first aspect or any corresponding embodiment thereof. Attached Figure Description

[0024] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0025] Figure 1 This is a schematic diagram of an application scenario according to an embodiment of the present invention; Figure 2 This is a schematic flowchart of a first type of digital anchor interaction method according to an embodiment of the present invention; Figure 3 This is a schematic diagram of a second process of the digital anchor interaction method according to an embodiment of the present invention; Figure 4 This is a structural block diagram of a digital anchor interactive device according to an embodiment of the present invention; Figure 5 This is a schematic diagram of the hardware structure of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0026] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0027] It is understood that before using the technical solutions disclosed in the various embodiments of the present invention, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in the present invention and their authorization should be obtained in accordance with relevant laws and regulations through appropriate means.

[0028] As an optional application scenario of this invention, such as Figure 1 As shown, application 101 is installed in terminal device 110, and user 130 can interact with application 101 through terminal device 110 and / or access device of terminal device 110.

[0029] For example, application 101 can be any application that provides question-and-answer related services. For instance, application 101 could be a question-and-answer interactive application, such as a text-to-text application, an image-to-text application, etc. Figure 1 In the application scenario shown, if application 101 is active, the terminal device 110 can display the interface 102 of application 101. The interface 102 may include various pages that application 101 can provide, such as interactive pages, settings pages, query pages, etc.

[0030] In some embodiments, terminal device 110 is communicatively connected to server 120 to provide services to application 101. Terminal device 110 may be a mobile terminal, fixed terminal, or portable terminal, etc., including but not limited to mobile phones, desktop computers, laptop computers, multimedia tablets, e-book devices, gaming devices, or any combination thereof, including accessories and peripherals of these devices or any combination thereof. In some embodiments, terminal device 110 may also support any type of interface, and server 120 may be various types of computing systems or servers capable of providing computing power, including but not limited to mainframes, edge computing nodes, computing devices in cloud environments, etc.

[0031] It should be noted that, Figure 1 This is merely an example of an application scenario and does not limit the scope of protection of this invention.

[0032] The embodiments of the present invention will now be described with reference to the accompanying drawings. It should be understood that the pages shown in the drawings are merely examples, and various page designs are possible in practice. The various graphic elements on the page may have different arrangements and different visual representations; one or more elements may be omitted or replaced, and one or more other elements may also be present, without any limitation in the embodiments of the present invention. Furthermore, the embodiments described below primarily pertain to terminal device 110. It should be understood that the actions described relative to terminal device 110 can be performed by application 101 on terminal device 110, or can be performed by application 101 in conjunction with its server (e.g., server 120).

[0033] This invention provides a digital anchor interaction method. By introducing deep learning technology, it achieves intelligent analysis and transformation of audience interaction information, and dynamically adjusts the digital anchor's state by combining historical data in a contextual memory bank, thereby generating more humanized and context-adaptive interactive content and actions. This method effectively solves the problems of mechanical interaction and lack of emotional expression in traditional digital anchor interactions, significantly improving the depth of interaction, audience participation, and dissemination effect of digital anchors.

[0034] According to an embodiment of the present invention, a digital anchor interaction method embodiment is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0035] This embodiment provides a digital anchor interaction method, which can be used in the aforementioned mobile terminals, such as mobile phones and tablets. Figure 2 This is a flowchart of a digital anchor interaction method according to an embodiment of the present invention, such as... Figure 2 As shown, the process includes the following steps: Step S201: Obtain audience interaction information and convert the interaction information into interaction events.

[0036] Taking the new-generation digital anchor as an example, the new-generation digital anchor refers to a virtual anchor image created and presented through digital technology in the agricultural field. Its main function is to conduct live interactive activities such as agricultural product marketing, agricultural knowledge popularization, and rural culture promotion. Interactive information refers to various forms of feedback generated by the audience during the live broadcast, such as text comments, voice chat, and virtual gift giving. Interactive events are structured data that drives changes in the state of the digital anchor after processing and abstracting interactive information. The contextual memory bank is a database that stores historical interactive data and the state of the digital anchor, providing important contextual information for the digital anchor's decision-making and state adjustment. In this embodiment, deep learning technology is mainly used for complex pattern recognition, semantic understanding, and sentiment analysis of audience interactive information, as well as for constructing and optimizing the state transition strategy model of the digital anchor.

[0037] In this embodiment, the first step is to acquire audience interaction information and convert it into interactive events. There are various ways to acquire audience interaction information. For example, text comments sent by viewers can be received in real time through the API interface provided by the live streaming platform. Alternatively, speech recognition technology can be used to convert the voice content of viewers during live chat into text information. Furthermore, virtual gifts sent by viewers can also be captured by the system as a form of interactive information. When converting this interactive information into interactive events, a preset rule matching method can be used. For example, when keywords such as "hello" or "thank you" are detected, they can be identified as greeting events or gratitude events. Alternatively, simple keyword extraction can be used to incorporate core words from audience comments as components of interactive events.

[0038] Step S202: Adjust the state of the digital anchor based on the interactive events, historical interactive events in the pre-stored context memory bank, and the historical state of the digital anchor.

[0039] Specifically, after acquiring and transforming interactive events, the state of the new farmer digital anchor needs to be adjusted based on the interactive event, historical interactive event data in the pre-stored contextual memory bank, and the historical state of the new farmer digital anchor. The contextual memory bank can be a simple database storing viewer comments and the digital anchor's responses over a past period. The historical state of the new farmer digital anchor can be a simple variable, such as recording whether the digital anchor is currently happy, calm, or confused. When adjusting the digital anchor's state, a judgment can be made based on the type of the current interactive event, combined with historical data and the current state, using preset conditional statements. For example, if the current interactive event is a question, and historical data shows that the digital anchor has demonstrated a professional attitude when answering similar questions, then the digital anchor's state can be adjusted to "professional."

[0040] Step S203: Based on the adjusted state of the digital anchor, generate and output the interactive content and actions of the digital anchor.

[0041] Specifically, based on the adjusted status of the new farmer digital anchor, the system will generate and synchronously output interactive content and actions for the anchor. Interactive content generation can use template filling; for example, based on the anchor's happy state, a reply can be randomly selected from a preset happy template library, such as "I'm happy to communicate with everyone." Action generation can select simple actions matching the current state from a preset action library; for example, when the anchor's state is happy, a smiling emoticon animation can be selected. Synchronized output of interactive content and actions can be achieved through simple timing control, ensuring that the verbal reply text and the corresponding emoticon actions are consistent in time.

[0042] Step S204: Update the interactive events, the adjusted digital anchor status, interactive content and actions, and the audience's feedback on the interactive content and actions to the pre-stored context memory bank.

[0043] Specifically, the interactive events, the adjusted status of the new farmer digital anchor, the interactive content and actions, and the audience's feedback on the interactive content and actions are updated to a pre-stored contextual memory bank. Updating the contextual memory bank can be done by appending records, adding all relevant data from this interaction as new entries to the database. Audience feedback on the interactive content and actions can include simple indicators such as whether viewers liked the post or continued to comment. This feedback information can serve as a reference for future adjustments to the digital anchor's status.

[0044] The digital broadcaster interaction method provided in this embodiment, based on deep learning, enables a more natural and emotionally profound interaction between the digital broadcaster and the audience through a series of closely linked steps. First, the system acquires and intelligently analyzes audience interaction information in real time, including voice messages, text comments, and virtual gifts, transforming them into structured interaction events. This process leverages the advantages of deep learning in speech recognition, natural language processing, and emotion recognition, ensuring accurate capture of audience intentions and emotions. Subsequently, based on the current interaction event and combined with rich contextual information such as historical interaction event data stored in a contextual memory bank in a time-series format, the digital broadcaster's historical behavior, emotion values, and focus points, the system dynamically adjusts the digital broadcaster's internal state in a smooth transition manner through preset state transition rules or strategy models. This state adjustment mechanism allows the digital broadcaster to simulate human emotional changes and psychological activities, making the interaction more coherent and human-like.

[0045] After the digital anchor's state is adjusted, the system generates a language response text that matches the current context based on the new state, and simultaneously generates corresponding facial expression animations, lip-sync animations, and body movements selected from a preset body movement library. This process not only ensures the semantic accuracy of the interactive content, but also allows the audience to clearly perceive the digital anchor's emotions and intentions through multimodal expression. For example, when the digital anchor is excited, the system generates a positive language response, accompanied by exaggerated facial expressions and large body movements, thereby enhancing the interactivity. To ensure the smoothness and naturalness of the output, the system also performs time-coordinated scheduling and emotional semantic consistency checks on the language response text, facial expression animations, lip-sync animations, and body movements, and adaptively adjusts the body movements according to the display area information of the current live broadcast screen to ensure that they are fully presented within the display area. Finally, the system drives the digital anchor model to synchronously output these adjusted interactive contents and movements, bringing an immersive interactive experience to the audience.

[0046] Furthermore, this embodiment of the invention emphasizes a continuous update mechanism for the contextual memory bank. After each interaction, the interactive event, the adjusted digital anchor status, the generated interactive content and actions, and the audience's feedback on these interactive content and actions (including comments, emotional inclinations, number of likes and gifts, and duration of interaction) are all promptly updated to the contextual memory bank. This continuous learning and iteration mechanism enables the digital anchor to constantly learn from historical interactions, optimize its behavioral patterns and emotional expression, thereby achieving self-evolution and continuously improving its interactive capabilities over time, ultimately achieving an interactive effect comparable to that of a real anchor.

[0047] The digital broadcaster interaction method proposed in this invention is a new farmer-based digital broadcaster interaction method based on deep learning, aiming to solve the pain points of existing digital broadcaster systems, such as mechanized interaction and lack of emotional expression. This invention constructs an intelligent interaction framework capable of perception, understanding, decision-making, and learning. Compared to traditional digital broadcasters based on preset scripts and simple response logic, this invention, by introducing deep learning technology, achieves deep semantic and sentiment analysis of audience interaction information, enabling digital broadcasters to more accurately understand audience intentions and emotions.

[0048] Specifically, related technologies, when handling audience interaction, are often limited to keyword matching or simple rule judgments, resulting in stiff and impersonal interactive content. However, this invention transforms audience voice messages, text comments, and virtual gifts into structured interactive events, and combines this with historical data from a contextual memory bank and the digital anchor's historical state to dynamically adjust the digital anchor's internal state. This state adjustment mechanism allows the digital anchor to simulate human emotional changes and psychological activities, thereby generating more context-adaptive and humanized interactive content and actions. For example, when an audience expresses curiosity about agricultural products, a traditional anchor might only provide a pre-set introduction; however, the digital anchor of this invention can adjust to an enthusiastic introduction state based on the audience's potential intentions and emotions, generating approachable language and facial expressions, significantly enhancing the depth of interaction and audience engagement.

[0049] Furthermore, the contextual memory bank update mechanism of this invention enables digital anchors to learn and evolve from each interaction. Viewer feedback on interactive content and actions, such as comments, emotional inclinations, number of likes and gifts, and dwell time, is recorded and used to optimize future interaction strategies. This contrasts sharply with related technologies where digital anchors lack continuous learning and self-optimization capabilities. Through this continuous iteration, the digital anchors of this invention can continuously improve the richness of their emotional expression and the intelligence level of their interaction strategies, ultimately achieving interactive effects comparable to real-life anchors, effectively addressing the shortcomings of existing digital anchors in establishing emotional connections and increasing fan loyalty.

[0050] This embodiment provides a digital anchor interaction method, which can be used in the aforementioned mobile terminals, such as mobile phones and tablets. Figure 3 This is a flowchart of a digital anchor interaction method according to an embodiment of the present invention, such as... Figure 3 As shown, the process includes the following steps: Step S301: Obtain audience interaction information and convert the interaction information into interaction events.

[0051] Specifically, step S301 includes: Step S3011: Collect and identify the audience's voice messages, text comments, and virtual gift information as audience interaction information; the virtual gift information includes the value and type of the virtual gift.

[0052] Specifically, interactive information refers to various types of data generated during audience interaction with digital livestreamers. This includes: voice chat (through voice calls) – real-time dialogue between viewers and livestreamers via voice link, which is typically collected using speech recognition technology to convert speech into processable text; text comments – comments, questions, or opinions expressed by viewers via text input in the livestream, which are usually collected by directly acquiring text data; and virtual gifts – virtual items purchased and gifted by viewers to livestreamers, which typically have different values ​​and symbolic meanings, and whose collection usually involves recording the type and quantity of gifts. By collecting and identifying this multimodal interactive information, a rich and comprehensive data foundation can be provided for the subsequent transformation of interactive events.

[0053] The aforementioned technical solutions enable comprehensive and detailed acquisition of audience interaction information, avoiding information gaps or biases that may result from a single form of interaction. This multimodal interactive information collection method significantly enhances the system's ability to perceive audience intentions and emotions, allowing new-generation digital anchors to more accurately understand audience needs and generate more targeted and humanized interactive content and actions, effectively improving the realism and immersion of the interactive experience.

[0054] Step S3012: Convert the live chat audio into live chat audio text, and perform semantic analysis and intent recognition on the live chat audio text and text comments to extract audience comment keywords and potential intents.

[0055] Converting live chat audio into text is typically achieved through speech recognition technology, aiming to transform the audience's spoken expressions into a computer-processable text format. Semantic analysis and intent recognition of live chat audio and text comments involve using Natural Language Processing (NLP) techniques to deeply understand the underlying meaning and potential purpose of audience comments. For example, it can identify whether the audience is asking a question, expressing agreement, offering suggestions, or complaining.

[0056] Step S3013: Based on the value and type of the virtual gift given, an emotion recognizer is used to identify the audience's emotions.

[0057] Step S3014: Integrate audience comment keywords, potential intentions, and audience emotions into interactive events.

[0058] During steps S3012 and S3014, core keywords and their underlying intentions are extracted from viewer comments. Furthermore, based on the value and type of virtual gifts given by viewers, an emotion recognizer can identify their emotional state. For example, giving a high-value rocket might express excitement or support, while giving flowers might express liking or gratitude. The emotion recognizer can be a pre-trained deep learning model used to analyze gift data and output corresponding emotion labels. Finally, these extracted viewer comment keywords, potential intentions, and viewer emotions are integrated to form a structured interaction event that comprehensively reflects the viewer's current interaction state and needs.

[0059] Through the above technical solutions, the embodiments of the present invention can more comprehensively and accurately understand the true intentions and emotional states of the audience. Specifically, by converting the voice chat into voice-text chat and performing semantic analysis and intent recognition on the voice-text chat and text comments, the keywords and potential intentions of the audience comments can be accurately extracted, avoiding the comprehension bias that may be caused by relying solely on a single text format. Simultaneously, identifying the audience's emotions based on the value and type of virtual gifts given further enriches the perception of the audience's emotional dimension, making the generation of interactive events more refined and humanized. Therefore, the generated interactive events can more accurately reflect the audience's interactive needs, providing high-quality input for subsequent adjustments to the state of new farmer digital anchors and the generation of interactive content, thereby significantly improving the intelligence and personalization level of digital anchor interaction and enhancing the user experience.

[0060] Step S302: Adjust the state of the digital anchor based on the interactive events, historical interactive events in the pre-stored context memory bank, and the historical state of the digital anchor.

[0061] Specifically, taking the new farmer digital anchor as an example, the adjustment of the new farmer digital anchor's status depends on the historical interaction event data in the pre-stored contextual memory bank. The aforementioned historical interaction event data can be understood as the historical interaction events within a preset time period stored in a time sequence, and the corresponding historical behaviors of the new farmer digital anchor.

[0062] Step S302 above includes: Step a: Based on the interactive events, historical interactive events, and the digital anchor's historical behavior, emotion value, and focus, calculate the updated state parameters of the digital anchor using preset state transition rules or strategy models, and adjust the digital anchor's current state to the state corresponding to the updated state parameters in a smooth transition manner.

[0063] The historical interaction event data refers to the collection of all audience interaction events that occurred within a specific time period during the live broadcast, as well as the responses of the new farmer digital anchors to these events, recorded and stored by the system. This data is designed to be stored in time series, meaning that each interaction event and its corresponding digital anchor behavior are assigned a timestamp and arranged in chronological order of occurrence.

[0064] The preset time period can be flexibly configured according to actual application needs. For example, it can be the most recent five minutes, ten minutes, half an hour or longer. The purpose is to ensure that the historical data used has sufficient timeliness and relevance, and to avoid using data that is too old or no longer applicable.

[0065] For example, historical interactive events refer to various interactive events that occurred when viewers interacted with the new farmer digital anchors in the past, such as viewers asking questions, commenting, liking, and sending gifts.

[0066] The historical behavior of new farmer digital anchors refers to the specific responses that new farmer digital anchors make to these historical interaction events, including but not limited to verbal replies, facial expressions, and body language.

[0067] Through the aforementioned technical solution, historical interaction event data is explicitly defined as historical interaction events within a preset time period stored in a time series, along with the corresponding historical behaviors of the new farmer digital anchors. This allows the system to obtain more in-depth and comprehensive contextual information when adjusting its state. This helps the system better understand the background of the current interaction event, predict the audience's potential intentions and emotional changes, and learn effective interaction strategies from the historical performance of the new farmer digital anchors. Therefore, the state adjustments of the new farmer digital anchors will be more consistent with actual interaction scenarios, avoiding the bias that may result from judging based solely on isolated events, and significantly improving the coherence, naturalness, and attractiveness of the new farmer digital anchors' interactions.

[0068] Specifically, the historical status of the aforementioned new-age digital anchors can be further defined in order to capture and reflect the internal dynamics of digital anchors in a more refined manner.

[0069] The historical status of new farmer digital anchors includes the emotional value used to quantify the emotional dimension of new farmer digital anchors within a preset time period, as well as the focus of new farmer digital anchors.

[0070] Among them, "emotional value" refers to a numerical value used to quantify the emotional dimension of new-generation digital anchors. It reflects the emotional tendency and intensity of the digital anchor at a specific point in time or within a preset time period. For example, emotional value can include multiple emotional dimensions such as happiness, sadness, surprise, and anger, and its degree is represented by numerical values. These emotional values ​​can be calculated and updated based on audience interaction information, historical interaction event data, and preset emotional models, aiming to provide a refined emotional basis for adjusting the digital anchor's state. Furthermore, "focus" refers to the specific topics, audience groups, or interactive content that new-generation digital anchors focus on within a preset time period. Focus can be understood as the digital anchor's attention focus, which can be dynamically adjusted based on the live broadcast content, audience feedback, and preset operational strategies. For example, when viewers frequently mention a certain agricultural product, that agricultural product can become the digital anchor's focus. The purpose is to guide the digital anchor's interaction direction, making their responses more targeted and relevant.

[0071] Through the aforementioned technical solutions, the process of digital anchors adjusting their state is imbued with deeper emotional and attentional considerations. This not only makes the digital anchors' interactive performances more natural and emotionally resonant but also significantly enhances the relevance and appeal of their interactions with viewers. By quantifying emotional values, digital anchors can better understand and respond to the emotional needs of their audience; by clearly defining their focus, they can more effectively guide the conversation and provide valuable information. Ultimately, this helps to build a more immersive and personalized live-streaming interactive experience, thereby increasing viewer engagement and satisfaction.

[0072] In some embodiments of the present invention described above, the state of the new farmer digital anchor is adjusted based on interactive events, historical interactive event data in the contextual memory bank, and the historical state of the new farmer digital anchor. However, if the state adjustment process lacks refined management, the state changes of the digital anchor may appear abrupt or unnatural, thereby affecting the audience's immersion and interactive experience.

[0073] In response, this embodiment of the invention further proposes the following steps for adjusting the state of a new farmer digital anchor based on interactive events, historical interactive event data in a pre-stored contextual memory bank, and the historical state of the new farmer digital anchor: based on interactive events, historical interactive events, and the historical behavior, emotion value, and focus of the new farmer digital anchor, the state of the new farmer digital anchor is calculated using a preset state transition rule or strategy model and adjusted in a smooth transition manner.

[0074] Specifically, the preset state transition rules or strategy model can be understood as a predefined set of logical judgments or a trained machine learning model. This rule or model is designed to comprehensively analyze the current interaction event, historical interaction events within a preset time period stored in the contextual memory bank in a time-series format, the corresponding historical behaviors of the new farmer digital anchor, and the new farmer digital anchor's historical states (including emotion values ​​used to quantify the new farmer digital anchor's emotional dimension and the new farmer digital anchor's focus), thereby calculating the new farmer digital anchor's next reasonable state. The strategy model can be a reinforcement learning model, a deep neural network model, or a rule-based expert system, whose goal is to predict and output the digital anchor state that best fits the current context and maintains consistency based on the input data.

[0075] The smooth transition method refers to avoiding sudden and drastic changes in the state of new-generation digital anchors. Instead, it involves a series of small, gradual adjustments to achieve the transition. For example, changes in emotional values ​​can be limited to a certain rate, or an interpolation algorithm can be used to generate an intermediate state between the old and new states. This makes the changes in the digital anchor's expressions, tone, and movements more natural and fluid, conforming to the gradual characteristics of human emotional expression.

[0076] Through the aforementioned technical solutions, the state adjustment process of digital anchors for new farmers has been given greater intelligence and naturalness. Digital anchors can transition between states in a more reasonable and coherent manner based on complex interactive scenarios, avoiding any sense of incongruity caused by sudden state changes. This not only enhances the realism and anthropomorphism of the digital anchors but also greatly improves the audience's immersion and satisfaction during the interaction process, making the human-computer interaction experience smoother and more emotionally resonant.

[0077] In some preferred embodiments, suppose a digital anchor, acting as a new-generation farmer, is live-streaming the introduction of agricultural products. Their initial emotional state is neutral to slightly positive, and their focus is on product features. When a viewer sends a comment with strong praise and encouragement (e.g., "The anchor is so good! The product looks amazing!") and gifts a high-value virtual gift, this interaction is identified as positive emotion and high intent. At this point, the strategy model comprehensively analyzes this positive interaction, the digital anchor's historical positive response behavior, the current emotional state, and the focus. Directly adjusting the emotional state to a very high positive level might appear unnatural. Therefore, the strategy model calculates a new, higher emotional state and smoothly transitions it. For example, over the next few seconds, it gradually increases the amplitude of the digital anchor's facial expression animation (smile), the speed and tone of their lip-sync animation, and selects a slow, gestured thank-you gesture from a pre-set body language library, rather than instantly switching to an exaggerated state of excitement. Simultaneously, the digital anchor's focus may smoothly shift from product features to gratitude and interaction with the viewer. This gradual adjustment makes the digital anchor's reactions more realistic and believable, much like a real anchor, effectively enhancing the interactive experience.

[0078] Step S303: Based on the adjusted state of the digital anchor, generate and output the interactive content and actions of the digital anchor.

[0079] Specifically, in the implementation process, if there is a lack of fine coordination and scheduling between interactive content, facial expressions, lip-syncing animation, and body movements, the interaction of the digital anchor may appear stiff and unnatural, making it difficult to effectively enhance the audience's immersion and interactive experience. To address this, this invention further proposes the steps described above for generating and synchronously outputting the interactive content and actions of the new farmer digital anchor.

[0080] Step S303 above includes: Step S3031: Generate a language response text that matches the adjusted state of the digital anchor, and generate facial expression animation and lip-sync animation corresponding to the language response text.

[0081] Specifically, generating language response text that matches the adjusted state of the new farmer digital anchor refers to generating emotionally charged and semantically coherent response text using a Natural Language Processing (NLG) model, based on the digital anchor's current mood, focus, and audience interaction events. For example, when the digital anchor is in a happy state, the generated text will contain positive vocabulary and tone. Simultaneously, generating facial expression animation and lip-syncing animation corresponding to the language response text involves using text-driven facial animation technology to generate real-time changes in the digital anchor's facial expressions and lip movements based on the semantic content, emotional tone, and pronunciation characteristics of the language response text, ensuring consistency between verbal and non-verbal expression. Facial expression animation can include various expressions such as smiling, surprise, and thinking, while lip-syncing animation accurately simulates the changes in mouth shape during pronunciation.

[0082] Step S3032: Select the corresponding limb movement from the preset limb movement library.

[0083] Furthermore, selecting appropriate body language from a pre-set body language library refers to intelligently choosing the most matching body language from a database containing various pre-set body language actions, based on the digital anchor's current state and the emotion and intent of their verbal responses. This body language library can include various actions such as nodding, waving, shrugging, and pointing, each of which may be associated with a specific emotion or semantic meaning. The selection process can be based on rule matching, machine learning models, or deep learning models.

[0084] Step S3033: Time-coordinated scheduling and synchronous output of the language response text, facial expression animation, lip-sync animation, and body movements.

[0085] Synchronizing and coordinating the output of spoken responses, facial animations, lip-syncing, and body language in a timely manner ensures that the digital anchor's language, facial expressions, lip movements, and body language are precisely aligned on the timeline and remain consistent in emotion and semantics. For example, when a digital anchor says "thank you" with a smile, their lip-syncing, facial animations, and body language (such as nodding or clasping their hands) should occur simultaneously and in harmony, creating a smooth and natural interactive experience. This coordination aims to eliminate potential delays or mismatches between different modalities, enhancing the overall realism of the interaction.

[0086] Through the aforementioned technical solutions, the generation and output of interactive content and actions for new-generation digital anchors have been given greater precision and coordination. This not only significantly enhances the naturalness and realism of the digital anchors' interactive performances, allowing viewers to experience a smoother and more emotionally resonant communication experience, but also strengthens viewer immersion and participation. Through the collaborative presentation of multimodal information, digital anchors can convey their emotions and intentions more accurately and vividly, effectively avoiding interaction barriers caused by inconsistencies between different forms of expression, thereby improving the overall interactive effect and user satisfaction.

[0087] In some preferred embodiments, suppose a viewer asks a digital broadcaster specializing in new-age agriculture, "How are these organic vegetables grown?"

[0088] First, based on the viewer's question and the digital anchor's current professional and helpful demeanor, the system generates a spoken response text, such as: "Hello! The cultivation of these organic vegetables mainly uses hydroponics technology, and the temperature and humidity of the growing environment are strictly controlled." At the same time, the system generates corresponding facial expression animations (such as focused, smiling) and lip-syncing animations based on this text to ensure synchronization with the speech.

[0089] Next, the system will select a gesture related to the explanation or guidance from a preset body gesture library. For example, the digital anchor may make a gesture as if indicating a certain planting process.

[0090] Finally, the system coordinates the generated spoken response text, facial expression animations, lip-syncing animations, and selected body movements in a timely manner. This means that when the digital anchor begins to speak, their lip movements will accurately synchronize with the pronunciation, their facial expressions will naturally switch according to the changes in the content being explained, and their body movements will appear at key information points and remain consistent with the spoken content. All these elements are output synchronously, forming a coherent and natural interactive scene, allowing viewers to receive information more intuitively and immersively.

[0091] In one optional implementation, the verbal response text, facial expression animation, lip-sync animation, and body movements are coordinated and output synchronously in time to realize the presentation of interactive content and actions of the new farmer digital anchor. The above step S3033 includes: Step b1: Obtain the display area information of the current live broadcast screen, and adjust the body movements according to the display area information so that the body movements are fully presented within the display area.

[0092] In practical applications, obtaining the display area information of the current live stream aims to clarify the visible range and layout of the digital anchor on the screen. This display area information can include the resolution of the live stream, the coordinate position and size of the digital anchor model within the image, and any potential obstructions. Based on this display area information, adaptive adjustments can be made to body movements. For example, if the digital anchor's preset body movements might extend beyond the screen boundaries, they can be scaled, panned, or cropped to ensure the movements are fully displayed within the display area, avoiding any negative impact on the viewer's experience due to incomplete movements.

[0093] Step b2 involves performing time synchronization and emotional semantic consistency checks on the language response text, facial expression animation, lip-sync animation, and body movements.

[0094] The timing synchronization of spoken responses, facial animations, lip-sync animations, and body language involves ensuring precise alignment of these different modalities on the timeline. For example, when a word in the spoken response is pronounced, the corresponding lip-sync animation should appear synchronously, the facial animation should match the emotion expressed by the word, and the relevant body language should occur at the appropriate time. Emotional semantic consistency checking involves comparing and verifying the emotions conveyed by the spoken response, the emotions expressed by the facial animation, and the emotions conveyed by the body language to ensure a high degree of consistency in emotional expression. Its purpose is to prevent digital anchors from speaking insincerely or displaying stiff expressions, thereby enhancing the realism and appeal of the interactive content.

[0095] Step b3 drives the digital anchor model to synchronously output the adjusted and consistent language response text, facial expression animation, lip-sync animation, and body movements.

[0096] Furthermore, driving the digital anchor model to synchronously output adjusted verbal responses, facial expression animations, lip-sync animations, and body movements refers to integrating all interactive elements—after undergoing the aforementioned adaptive adjustments, time synchronization, and emotional semantic consistency checks—through the digital anchor's rendering engine or animation system, and presenting them to the audience in a unified and coordinated manner in real time. This typically involves complex real-time rendering techniques and animation playback control mechanisms to ensure that all elements achieve optimal synchronization and expressiveness both visually and aurally.

[0097] Through the aforementioned technical solutions, the interactive content and actions of digital anchors can better adapt to the live streaming environment during output, avoiding visual problems caused by body movements exceeding the frame or conflicting with frame elements. Furthermore, rigorous time synchronization and emotional semantic consistency checks significantly enhance the naturalness, realism, and engaging nature of digital anchor interactions, providing viewers with a more immersive and enjoyable interactive experience, effectively strengthening the interactive effect and user engagement of digital anchors.

[0098] Step S304: Update the interactive events, the adjusted digital anchor status, interactive content and actions, and the audience's feedback on the interactive content and actions to the pre-stored context memory bank.

[0099] Among these factors, audience feedback on interactive content and actions is crucial for updating the contextual memory bank. This feedback includes audience comments and emotional inclinations, the number of likes and gifts given, and the duration of audience engagement.

[0100] Viewer comments refer to the opinions, questions, or feelings expressed by viewers through text or voice during the live streams hosted by digital farmers. Sentiment analysis involves using natural language processing technology to analyze viewer comments and identify whether the emotions expressed are positive, negative, or neutral. For example, sentiment analysis models can be used to score comment texts, thereby quantifying the viewer's sentiment.

[0101] The number of likes and gifts refers to the degree to which viewers express their approval and support for the interactive content and actions of the new farmer digital anchors by clicking the "like" button or sending virtual gifts. The number of likes and gifts can directly reflect the viewers' satisfaction and engagement.

[0102] Viewer dwell time refers to the length of time a viewer continuously watches a live stream. A longer dwell time usually indicates that the viewer has a higher level of interest and engagement with the live stream content.

[0103] The above technical solutions can ensure that the feedback data stored in the contextual memory bank has higher dimensions and stronger representativeness.

[0104] Specifically, by integrating audience comments and sentiment, a deeper understanding of audience opinions and emotional responses to specific interactive content can be achieved; by counting the number of likes and gifts, audience satisfaction and support can be quantified; and by monitoring audience dwell time, the attractiveness of the live stream content and the depth of audience engagement can be assessed. Thus, the contextual memory bank can obtain a more comprehensive and refined profile of audience feedback, providing a more accurate basis for adjusting the performance of new-generation digital broadcasters, thereby improving the interactive effects and user experience of these broadcasters.

[0105] The digital anchor interaction method provided in this embodiment acquires audience interaction information and transforms it into interactive events. It then combines historical data from a contextual memory bank with the digital anchor's historical states to adjust its current state, thereby generating and synchronously outputting interactive content with emotions and actions. Feedback information from the entire interaction process is updated in the contextual memory bank. This method effectively solves the problems of existing digital anchor systems, such as mechanical interaction, lack of emotional expression, and in-depth psychological state presentation.

[0106] This embodiment introduces a contextual memory bank and a state adjustment mechanism, enabling digital anchors to dynamically adjust their states based on real-time interactive contexts and historical experience. This makes their interactive behavior more coherent, human-like, and emotional, overcoming the shortcomings of existing technologies such as stiff interaction and emotional emptiness. Accordingly, this embodiment of the invention can significantly enhance the emotional connection and interaction depth between digital anchors and viewers, strengthen viewer participation and engagement, and meet the urgent demand for intelligent and emotionally engaging digital anchors in the new business landscape.

[0107] This embodiment also provides a digital broadcaster interaction device, which is used to implement the above embodiments and preferred embodiments; details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that implements a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0108] This embodiment provides a digital broadcaster interactive device, such as... Figure 4 As shown, it includes: The information acquisition and conversion module 401 is used to acquire audience interaction information and convert the interaction information into interactive events.

[0109] The state adjustment module 402 is used to adjust the state of the digital anchor based on interactive events, historical interactive events in the pre-stored context memory bank, and the historical state of the digital anchor.

[0110] The interactive content and action generation module 403 is used to generate and output the interactive content and actions of the digital anchor based on the adjusted state of the digital anchor.

[0111] The context memory bank update module 404 is used to update the pre-stored context memory bank with interactive events, the adjusted status of the digital anchor, interactive content and actions, and audience feedback information on interactive content and actions.

[0112] In some optional implementations, the information acquisition and conversion module 401 includes: The information collection unit is used to collect and identify the audience's voice messages, text comments, and virtual gift information as audience interaction information; the virtual gift information includes the value and type of the virtual gift.

[0113] In some optional implementations, the information acquisition and conversion module 401 further includes: The semantic analysis and intent recognition unit is used to convert live chat audio into live chat audio text, and to perform semantic analysis and intent recognition on the live chat audio text and text comments to extract audience comment keywords and potential intents.

[0114] The audience emotion recognition unit is used to identify audience emotions based on the value and type of virtual gifts given, using an emotion recognizer.

[0115] The conversion unit is used to integrate audience comment keywords, potential intentions, and audience emotions into interactive events.

[0116] In some optional implementations, historical interaction events include: historical interaction events stored in a time sequence within a preset time period and the corresponding historical behaviors of the digital anchor; the historical state of the digital anchor includes: emotional values ​​used to quantify the emotional dimension of the digital anchor within the preset time period and the digital anchor's focus; the state adjustment module 402 includes: The parameter update and state adjustment unit is used to calculate the updated state parameters of the digital anchor based on interactive events, historical interactive events, the historical behavior and emotion value of the digital anchor, and the focus of the digital anchor, through preset state transition rules or strategy models, and adjust the current state of the digital anchor to the state corresponding to the updated state parameters in a smooth transition manner.

[0117] In some optional implementations, the interactive content and action generation module 403 includes: The interactive content generation unit is used to generate language response text that matches the adjusted digital anchor's state, and to generate facial expression animations and lip-sync animations corresponding to the language response text.

[0118] The limb movement selection unit is used to select the corresponding limb movement from the preset limb movement library.

[0119] The collaborative scheduling and output unit is used to perform time-coordinated scheduling and synchronous output of language response text, facial expression animation, lip-sync animation, and body movements.

[0120] In some optional implementations, the cooperative scheduling and output unit includes: The motion adjustment subunit is used to obtain the display area information of the current live broadcast screen and adjust the body movements according to the display area information so that the body movements are fully presented within the display area.

[0121] The consistency check subunit is used to perform time synchronization and emotional semantic consistency checks on language response text, facial expression animation and lip-sync animation, and body movements.

[0122] The output subunit is used to drive the digital anchor model to synchronously output adjusted and consistent language response text, facial expression animation, lip-sync animation, and body movements.

[0123] In some alternative implementations, audience feedback on interactive content and actions includes: audience comments and sentiment, audience likes, the number of gifts given, and the duration of audience engagement.

[0124] The digital anchor interaction device provided in this embodiment of the invention can execute the digital anchor interaction method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects for executing the method. Further functional descriptions of the various modules and units described above are the same as in the corresponding embodiments described above, and will not be repeated here.

[0125] Figure 5 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention.

[0126] The following is a detailed reference. Figure 5 The diagram illustrates a structural schematic suitable for implementing an electronic device according to embodiments of the present invention. The electronic device may include a processor (e.g., a central processing unit, graphics processor, etc.) 501, which can perform various appropriate actions and processes according to a program stored in read-only memory (ROM) 502 or a program loaded from memory 508 into random access memory (RAM) 503. The RAM 503 also stores various programs and data required for the operation of the electronic device. The processor 501, ROM 502, and RAM 503 are interconnected via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0127] Typically, the following devices can be connected to I / O interface 505: input devices 506 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 507 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; memory devices 508 including, for example, magnetic tapes, hard disks, etc.; and communication devices 509. Communication device 509 allows electronic devices to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 5 Electronic devices with various devices are shown, but it should be understood that it is not required to implement or have all of the devices shown, and more or fewer devices may be implemented or have instead.

[0128] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 509, or installed from a memory 508, or installed from a ROM 502. When the computer program is executed by the processor 501, it performs the functions defined in the digital broadcaster interaction method of the embodiments of the present invention.

[0129] Figure 5 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments of the present invention.

[0130] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code. When the software or computer code is accessed and executed by the computer, processor, or hardware, the digital broadcaster interaction method shown in the above embodiments is implemented.

[0131] A portion of this invention can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide the methods and / or technical solutions according to the invention through the operation of the computer. Those skilled in the art will understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executing the instructions, or the computer compiling the instructions and then executing the corresponding compiled program, or the computer reading and executing the instructions, or the computer reading and installing the instructions and then executing the corresponding installed program. Here, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to a computer.

[0132] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A digital anchor interaction method, characterized in that, The method includes: Acquire audience interaction information and convert the interaction information into interactive events; The state of the digital anchor is adjusted based on the interactive events, historical interactive events in the pre-stored context memory bank, and the historical state of the digital anchor. Based on the adjusted status of the digital anchor, generate and output the interactive content and actions of the digital anchor; The interactive events, the adjusted digital anchor status, the interactive content and actions, and the audience's feedback on the interactive content and actions are updated to the pre-stored context memory bank.

2. The method according to claim 1, characterized in that, The acquisition of audience interaction information includes: Collect and identify the audience's voice messages, text comments, and virtual gift information as audience interaction information; the virtual gift information includes the value and type of the virtual gift.

3. The method according to claim 2, characterized in that, Converting the interactive information into interactive events includes: The live chat audio is converted into live chat audio text, and semantic analysis and intent recognition are performed on the live chat audio text and the text comments to extract audience comment keywords and potential intents; Based on the value and type of the virtual gifts given, an emotion recognition device is used to identify the audience's emotions. The audience comments, keywords, underlying intentions, and emotions are integrated and transformed into interactive events.

4. The method according to claim 1, characterized in that, The historical interaction events include: historical interaction events stored in a time sequence within a preset time period and the corresponding historical behaviors of the digital anchor; the historical state of the digital anchor includes: the emotional value used to quantify the emotional dimension of the digital anchor and the focus of the digital anchor within the preset time period. Based on the interactive events, historical interactive events in the pre-stored context memory bank, and the digital anchor's historical state, adjust the digital anchor's state, including: Based on the interactive events, the historical interactive events and the corresponding historical behaviors of the digital anchor, the emotion value and the focus of the digital anchor, the updated state parameters of the digital anchor are calculated through preset state transition rules or strategy models, and the current state of the digital anchor is adjusted to the state corresponding to the updated state parameters in a smooth transition manner.

5. The method according to claim 1, characterized in that, The process of generating and outputting interactive content and actions of the digital anchor based on the adjusted state of the digital anchor includes: Generate a language response text that matches the adjusted state of the digital anchor, and generate facial expression animation and lip-sync animation corresponding to the language response text; Select the appropriate body movement from the preset body movement library; The language response text, facial expression animation, lip-sync animation, and body movements are time-coordinated and output synchronously.

6. The method according to claim 5, characterized in that, The time-coordinated scheduling and synchronous output of the language response text, facial expression animation, lip-sync animation, and body movements includes: Obtain the display area information of the current live broadcast screen, and adjust the body movements according to the display area information so that the body movements are fully presented within the display area; Perform time synchronization and emotional semantic consistency checks on the language response text, the facial expression animation and lip-sync animation, and the body movements; The digital anchor model synchronously outputs the adjusted and consistent language response text, facial expression animation, lip-sync animation, and body movements after the consistency check is passed.

7. The method according to claim 1, characterized in that, The audience's feedback information on the interactive content and actions includes: audience comments and emotional inclinations, audience likes, the number of gifts given, and the audience's dwell time.

8. A digital broadcaster interactive device, characterized in that, The device includes: The information acquisition and conversion module is used to acquire audience interaction information and convert the interaction information into interactive events; The state adjustment module is used to adjust the state of the digital anchor based on the interactive events, historical interactive events in the pre-stored context memory bank, and the historical state of the digital anchor. The interactive content and action generation module is used to generate and output the interactive content and actions of the digital anchor based on the adjusted state of the digital anchor. The context memory bank update module is used to update the interactive events, the adjusted state of the digital anchor, the interactive content and actions, and the audience's feedback information on the interactive content and actions to the pre-stored context memory bank.

9. An electronic device, characterized in that, include: A memory and a processor are communicatively connected, the memory storing computer instructions, and the processor executing the computer instructions to perform the digital anchor interaction method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing the computer to perform the digital anchor interaction method according to any one of claims 1 to 7.

Images