Human-computer collaborative creation method, apparatus, terminal device, and storage medium

Abstract

The invention provides a man-machine collaborative creation method and device, terminal equipment and a storage medium, and is suitable for the technical field of man-machine interaction. The method comprises the following steps: acquiring physiological feature information of a target object, and determining emotional state information of the target object based on the physiological feature information; the resonance direction of the target object is determined based on the emotional state information, the resonance direction comprises forward resonance and reverse resonance, the forward resonance is used for representing that the emotional state of the target object is positive emotional, and the reverse resonance is used for representing that the emotional state of the target object is negative emotional; based on the resonance direction, acquiring an emotion style template corresponding to the resonance direction from a preset suggestion generation strategy library; and determining creation suggestion content based on the emotion style template. According to the embodiment of the invention, the corresponding creation suggestion content can be given based on the emotion of the user, and the man-machine cooperation efficiency is improved.

Description

(19) State Intellectual Property Office (12) Invention Patent Application (10) Application Publication Number (43) Application Publication Date (21) Application Number 202511096466.0 (22) Application Date 2025.08.04 (71) Applicant Sun Yuqian Address No. 1, Scenic Road, Huangwan Township, Emeishan City, Leshan City, Sichuan Province, 614200 (72) Inventor Sun Yuqian (74) Patent Agency Shenzhen Zhongyi United Intellectual Property Agency Co., Ltd. 44414 Patent Attorney Hu Juan (51) Int.Cl. G06F 18 / 24 (2023.01) G06F 18 / 22 (2023.01) G06F 16 / 335 (2019.01) G06F 16 / 338 (2019.01) (54) Invention Title: Human-Computer Collaborative Creation Method, Apparatus, Terminal Device, and Storage Medium (57) Abstract: This application provides a human-computer collaborative creation method, apparatus, terminal device, and storage medium, applicable to the field of human-computer interaction technology. The method includes: acquiring physiological characteristic information of a target object, and determining the emotional state information of the target object based on the physiological characteristic information; determining the resonance direction of the target object based on the emotional state information: wherein the resonance direction includes positive resonance and negative resonance, positive resonance is used to indicate that the emotional state of the target object is positive, and negative resonance is used to indicate that the emotional state of the target object is negative; based on the resonance direction, acquiring an emotional style template corresponding to the resonance direction from a preset suggestion generation strategy library; and determining the creation suggestion content based on the emotional style template. The embodiments of this application can provide corresponding creation suggestion content based on the user's emotions, improving the efficiency of human-computer collaboration. Claims 2 pages, Description 18 pages, Drawings 4 pages, CN 121093034 A 2025.12.09 CN 1 21 09 30 34 A 1. A human-computer collaborative creation method, characterized in that it includes: acquiring physiological characteristic information of a target object, and determining the emotional state information of the target object based on the physiological characteristic information; determining the resonance direction of the target object based on the emotional state information: wherein the resonance direction includes positive resonance and negative resonance, the positive resonance is used to indicate that the emotional state of the target object is a positive emotion, and the negative resonance is used to indicate that the emotional state of the target object is a negative emotion; acquiring an emotional style template corresponding to the resonance direction from a preset suggestion generation strategy library based on the resonance direction; determining the creation suggestion content based on the emotional style template. 2. The human-computer collaborative creation method according to claim 1, characterized in that, after determining the creation suggestion content based on the emotional style template, it further includes: acquiring feedback information from the target object regarding the creation suggestion content; wherein the feedback information includes the followingAt least one of the following: response behavior information and reading status information, wherein the response behavior information includes any one of adopting the creative suggestion content, skipping the creative suggestion content, or modifying the creative suggestion content; and the reading status information includes at least one of the speech signal and facial expression features of the target object when reading the creative suggestion content. Based on the feedback information, the call weight of the emotional style template is adjusted; wherein the call weight is used to represent the probability that the emotional style template is selected. 3. The human-computer collaborative creation method according to claim 2, wherein adjusting the call weight of the emotional style template based on the feedback information includes: inputting the emotional style template and the feedback information into a resonance enhancement model to obtain a fit score; wherein the resonance enhancement model is trained based on training samples, and the training samples include the mapping relationship between the emotional style template and the feedback information and the corresponding fit score; and adjusting the call weight of the emotional style template based on the fit score. 4. The human-computer collaborative creation method according to claim 1, characterized in that the physiological feature information includes at least one of the physiological signals and behavioral signals of the target object collected during the human-computer collaborative creation process, wherein the physiological signals include at least one of electroencephalogram (EEG), heart rate, and skin conductance response, and the behavioral signals include at least one of the speech signal and facial expression features of the target object when reading the current creative content; the step of determining the emotional state information of the target object based on the physiological feature information includes: inputting the physiological feature information into a fusion processing module for signal preprocessing to obtain data to be processed; wherein the fusion method of the fusion processing module includes at least one of feature-level splicing fusion, attention-based weighted fusion, and gating-based dynamic fusion; inputting the data to be processed into an emotion recognition model to obtain a multi-dimensional emotion state vector as the emotion state information; wherein the emotion recognition model is obtained based on a support vector machine, a long short-term memory network, or an ensemble classification model. 5. The human-computer collaborative creation method according to claim 4, characterized in that, determining the resonance direction of the target object based on the emotional state information includes: if the multidimensional emotional state vector belongs to a preset positive resonance interval, then the resonance direction of the target object is determined to be positive resonance; if the multidimensional emotional state vector belongs to a preset negative resonance interval, then the resonance direction of the target object is determined to be negative resonance. Claims 1 / 2 Page 2 CN 121093034 A 6. The human-computer collaborative creation method according to any one of claims 1-5, characterized in that, obtaining the emotional style template corresponding to the resonance direction from a preset suggestion generation strategy library based on the resonance direction includes:The method involves: acquiring current creative content and extracting contextual tags from the current creative content; wherein the contextual tags include at least one of scene keywords, emotional themes, and plot nodes; and acquiring an emotional style template corresponding to the resonance direction from a preset suggestion generation strategy library based on the resonance direction and the contextual tags. 7. The human-computer collaborative creation method according to any one of claims 1-5, wherein the emotional style template includes at least one of semantic features, intonation style, pragmatic rhythm, and sentence structure; and determining the creative suggestion content based on the emotional style template, comprising: generating the creative suggestion content based on the emotional style template from at least one dimension of language style, creative structure, and emotional resonance. 8. A human-computer collaborative creation device, characterized in that it comprises: a first determining module, configured to acquire physiological characteristic information of a target object, and determine the emotional state information of the target object based on the physiological characteristic information; a second determining module, configured to determine the resonance direction of the target object based on the emotional state information: wherein the resonance direction includes positive resonance and negative resonance, the positive resonance being used to indicate that the emotional state of the target object is a positive emotion, and the negative resonance being used to indicate that the emotional state of the target object is a negative emotion; a template acquisition module, configured to acquire an emotional style template corresponding to the resonance direction from a preset suggestion generation strategy library based on the resonance direction; and a third determining module, configured to determine the creation suggestion content based on the emotional style template. 9. A terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor, when executing the computer program, implements the steps of the method as described in any one of claims 1 to 7. 10. A computer-readable storage medium storing a computer program, characterized in that the computer program, when executed by a processor, implements the steps of the method as described in any one of claims 1 to 7. Claims 2 / 2 Page 3 CN 121093034 A Human-Computer Collaborative Creation Method, Apparatus, Terminal Equipment and Storage Medium Technical Field

[0001] This application relates to the field of human-computer interaction technology, and in particular to a human-computer collaborative creation method, apparatus, terminal equipment and storage medium. Background Art

[0002] Currently, most artificial intelligence creation assistance methods on the market are based on language models, and their generation process mostly adopts a passive response mechanism, that is, the user inputs content or instructions, and a response is generated according to the probability of language prediction.

[0003] In recent years, some language model systems have introduced the expression structure of "suggestive tone", such as "You can try..." "I suggest...", etc., to enhance the humanized interactive experience. However, such suggestive tone is usually templated language or context.The predicted results do not possess the ability to truly understand user intent. In other words, the so-called "suggestions" are merely simulations in linguistic form. Such "pseudo-suggestions" may cause users to misjudge the system's initiative in practical applications, or even form incorrect perceptions, leading to low efficiency in human-computer collaboration. Summary of the Invention

[0004] In view of this, embodiments of this application provide a human-computer collaborative creation method, apparatus, terminal device, and storage medium to solve the problem of low efficiency in human-computer collaboration in the prior art.

[0005] A first aspect of embodiments of this application provides a human-computer collaborative creation method, comprising:

[0006] obtaining physiological characteristic information of a target object, and determining the emotional state information of the target object based on the physiological characteristic information;

[0007] determining the resonance direction of the target object based on the emotional state information: wherein the resonance direction includes positive resonance and negative resonance, positive resonance is used to indicate that the emotional state of the target object is positive, and negative resonance is used to indicate that the emotional state of the target object is negative;

[0008] obtaining an emotional style template corresponding to the resonance direction from a preset suggestion generation strategy library based on the resonance direction;

[0009] determining the content of the creation suggestion based on the emotional style template.

[0010] In one possible implementation, after determining the content of the creative suggestions based on the emotional style template, the method further includes:

[0011] Obtaining feedback information from the target object regarding the content of the creative suggestions; wherein the feedback information includes at least one of the following: response behavior information and reading status information, wherein the response behavior information includes any one of adopting the content of the creative suggestions, skipping the content of the creative suggestions, and modifying the content of the creative suggestions, and the reading status information includes at least one of the voice signal and facial expression features of the target object when reading the content of the creative suggestions;

[0012] Adjusting the call weight of the emotional style template based on the feedback information; wherein the call weight is used to represent the probability of the emotional style template being selected and obtained.

[0013] In one possible implementation, the call weight of the emotional style template is adjusted based on feedback information, including:

[0014] Inputting the emotional style template and feedback information into a resonance enhancement model to obtain a fit score; wherein, the resonance enhancement model is trained based on training samples, and the training samples include the mapping relationship between the emotional style template and feedback information and the fit score corresponding to the specification page 1 / 18 4 CN 121093034 A;

[0015] Adjusting the call weight of the emotional style template based on the fit score.

[0016] In one possible implementation, the physiological feature information includes at least one of the physiological signals and behavioral signals of the target object collected during the human-computer collaborative creation process, the physiological signals including at least one of electroencephalogram (EEG), heart rate, and skin conductance response, and the behavioral signals including at least one of the speech signal and facial expression features of the target object when reading the current creative content;

[0017] Based on physiological feature information, the emotional state information of the target object is determined, including:

[0018] Inputting the physiological feature information into the fusion processing module for signal preprocessing to obtain the data to be processed; wherein, the fusion method of the fusion processing module includes at least one of feature-level splicing fusion, weighted fusion based on attention mechanism, and dynamic fusion based on gating mechanism;

[0019] Inputting the data to be processed into the emotion recognition model to obtain a multi-dimensional emotion state vector as the emotion state information; wherein, the emotion recognition model is obtained based on support vector machine or long short-term memory network or ensemble classification model.

[0020] In a possible implementation, based on the emotional state information, the resonance direction of the target object is determined, including:

[0021] If the multi-dimensional emotion state vector belongs to a preset positive resonance interval, the resonance direction of the target object is determined to be positive resonance;

[0022] If the multi-dimensional emotion state vector belongs to a preset negative resonance interval, the resonance direction of the target object is determined to be negative resonance.

[0023] In one possible implementation, based on the resonance direction, an emotional style template corresponding to the resonance direction is obtained from a preset suggestion generation strategy library, including:

[0024] Obtaining the current creative content and extracting contextual tags from the current creative content; wherein, the contextual tags include at least one of scene keywords, emotional themes, and plot nodes;

[0025] Based on the resonance direction and contextual tags, an emotional style template corresponding to the resonance direction is obtained from a preset suggestion generation strategy library.

[0026] In one possible implementation, the emotional style template includes at least one of semantic features, intonation style, pragmatic rhythm, and sentence structure;

[0027] Based on the emotional style template, the creative suggestion content is determined, including:

[0028] Based on the emotional style template, the creative suggestion content is generated from at least one dimension of language style dimension, creative structure dimension, and emotional resonance dimension.

[0029] A second aspect of the present application provides a human-computer collaborative creation device, comprising:

[0030] a first determining module, configured to acquire physiological characteristic information of a target object, and determine the emotional state information of the target object based on the physiological characteristic information;

[0031] a second determining module, configured to determine the resonance direction of the target object based on the emotional state information: wherein the resonance direction includes positive resonance and negative resonance, positive resonance is used to indicate that the emotional state of the target object is a positive emotion, and negative resonance is used to indicate that the emotional state of the target object is a negative emotion;

[0032] a template acquisition module, configured to acquire an emotional style template corresponding to the resonance direction from a preset suggestion generation strategy library based on the resonance direction;

[0033] a third determining module, configured to determine the creation suggestion content based on the emotional style template.

[0034] A third aspect of the present application provides a terminal device, comprising a memory, a processor, and a storage device.A computer program stored in a storage medium and executable on a processor, wherein when the processor executes the computer program, it implements the steps of the method as described in the first aspect of the specification (page 2 / 18, CN 121093034 A).

[0035] A fourth aspect of the present application provides a computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, it implements the steps of the method as described in the first aspect.

[0036] The beneficial effects of the present application embodiments compared with the prior art are:

[0037] The human-computer collaborative creation method of the first aspect of the present application embodiments can obtain physiological characteristic information of the target object, and determine the emotional state information of the target object based on the physiological characteristic information, then determine the resonance direction of the target object based on the emotional state information, and then obtain the emotional style template corresponding to the resonance direction from a preset suggestion generation strategy library based on the resonance direction, and then determine the creation suggestion content based on the emotional style template. Therefore, the present application embodiments can dynamically incorporate the user's real cognitive state and emotional feedback into the creation suggestion content generation mechanism, and can give corresponding creation suggestion content based on the user's emotions, improve the efficiency of human-computer collaboration, and thus improve the user experience.

[0038] It is understood that the beneficial effects of the second to fourth aspects described above can be referred to the relevant description in the first aspect described above, and will not be repeated here. Brief Description of the Drawings

[0039] In order to more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0040] FIG1 is a flowchart of a human-computer collaborative creation method provided by an embodiment of this application;

[0041] FIG2 is a flowchart of another human-computer collaborative creation method provided by an embodiment of this application;

[0042] FIG3 is a flowchart of obtaining an emotional style template corresponding to the resonance direction provided by an embodiment of this application;

[0043] FIG4 is a structural schematic diagram of a human-computer collaborative creation system provided by an embodiment of this application;

[0044] FIG5 is a structural schematic diagram of a human-computer collaborative creation device provided by an embodiment of this application;

[0045] FIG6 is a structural schematic diagram of a terminal device provided by an embodiment of this application. Detailed Description of Embodiments

[0046] In the following description, specific details such as particular system structures and techniques are set forth for illustrative purposes and not for limiting purposes, so as to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods are omitted so as not to obscure the description of this application with unnecessary details.

[0047] It should be understood that, when used in this specification and the appended claims, the term "comprising" indicates the presence of the described feature, integral, step, operation, element, and / or component, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components, and / or collections thereof.

[0048] It should also be understood that, as used in this specification and the appended claims, the term "and / or" refers to any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0049] As used in this specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrase "if determined" or "if detected [the described condition or event]" may be interpreted, depending on the context, as meaning "once determined," "in response to determination," "once detected [the described condition or event]," or "in response to detection [the described condition or event]." Specification 3 / 18 pages 6 CN 121093034 A

[0050] In addition, in the description of this application specification and the appended claims, the terms "first", "second", "third", etc. are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0051] References such as "one embodiment" or "some embodiments" described in this application specification mean that one or more embodiments of this application include a specific feature, structure or characteristic described in connection with that embodiment. Thus, the phrases "in one embodiment", "in some embodiments", "in other embodiments", "in other embodiments", etc., appearing in different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized. The terms "comprising", "including", "having", and their variations all mean "including but not limited to", unless otherwise specifically emphasized.

[0052] The technical solutions of this application and how the technical solutions of this application solve the above-mentioned technical problems are described in detail below with specific embodiments. It should be noted that the following embodiments can be referenced, borrowed, or combined with each other. Identical terms, similar features, and similar implementation steps in different embodiments will not be described again.

[0053] Referring to Figure 1, this application embodiment provides a flowchart of a human-computer collaborative creation method. As shown in Figure 1, the human-computer collaborative creation method of this application embodiment includes steps S101 to S104.

[0054] S101: Obtain physiological characteristic information of the target object, and determine the emotional state information of the target object based on the physiological characteristic information.

[0055] In some embodiments, physiological feature information includes at least one of physiological signals and behavioral signals of the target object collected during human-computer collaborative creation. The physiological signals include at least one of electroencephalogram (EEG), heart rate, and skin conductance response. The behavioral signals include at least one of speech signals and facial expression features of the target object when reading the current creative content.

[0056] Based on the physiological feature information, determining the emotional state information of the target object includes:

[0057] Inputting the physiological feature information into a fusion processing module for signal preprocessing to obtain data to be processed; wherein, the fusion method of the fusion processing module includes at least one of feature-level splicing fusion, weighted fusion based on attention mechanism, and dynamic fusion based on gating mechanism;

[0058] Inputting the data to be processed into an emotion recognition model to obtain a multi-dimensional emotion state vector as emotion state information; wherein, the emotion recognition model is obtained based on a support vector machine, a long short-term memory network, or an ensemble classification model.

[0059] In this embodiment of the application, multimodal data is collected throughout the entire creation process, divided into two stages:

[0060] (1) During the AI ​​collaborative creation process, the following signals are collected:

[0061] Electroencephalogram (EEG): α / β power changes are collected;

[0062] Heart rate (PPG): HRV and pulse rhythm are collected;

[0063] Skin conductance (GSR): SCR frequency and peak value are collected;

[0064] Facial expression features: Action unit (AU) and expression intensity are collected in real time through a camera to supplement physiological data and assist in the identification of immediate emotional state.

[0065] Wherein, EEG (Electroencephalogram) represents electroencephalogram; PPG (Photoplethysmography) represents photoplethysmography; GSR (Galvanic Skin Response) represents skin conductance response.

[0066] (2) After the creation is completed, the user reads the AI-generated work and collects the following signals:

[0067] Facial expression features: collect facial expression changes during the reading process to evaluate the "acceptance / resistance" type of feedback emotion;

[0068] Speech signal: extract FO fundamental frequency, speech rate, sound intensity, and tone fluctuation to identify psychological attitude (such as hesitation, affirmation, negation) and content resonance degree.

[0069] The emotion recognition in this application embodiment is to collect EEG, PPG, GSR, and facial expression signals in real time during the creation process; and to evaluate the "feedback attitude" again through speech and facial expression during the reading stage, and to integrate the multimodal data to obtain a multi-dimensional emotional state vector E_user_t.

[0070] S102. Based on the emotional state information, determine the resonance direction of the target object: wherein, the resonance direction includes positive resonance.Resonance and reverberation, positive resonance is used to indicate that the target object's emotional state is positive, and reverberation is used to indicate that the target object's emotional state is negative.

[0071] In some embodiments, based on emotional state information, determining the resonance direction of the target object includes:

[0072] If the multidimensional emotional state vector belongs to a preset positive resonance interval, then the resonance direction of the target object is determined to be positive resonance;

[0073] If the multidimensional emotional state vector belongs to a preset negative resonance interval, then the resonance direction of the target object is determined to be negative resonance.

[0074] This application embodiment can integrate the above-mentioned multiple signals, extract features and form a unified multidimensional emotional state vector E_user_t, the multidimensional emotional state vector E_user_t contains timestamps of "immediate state" (in creation) and "feedback emotion" (reading stage); retain the later acquisition of facial and voice as "supplementary resonance markers".

[0075] The resonance direction judgment in this application embodiment is: if E_user_t∈positive interval, the "positive resonance strategy" is activated accordingly; if E_user_t∈negative interval, the "reverberation strategy" is activated accordingly.

[0076] S103. Based on the resonance direction, obtain the emotional style template corresponding to the resonance direction from the preset suggestion generation strategy library.

[0077] The positive resonance strategy and the negative resonance strategy of the suggestion generation strategy library in this application embodiment correspond to the following two mechanisms:

[0078] 1. Positive resonance mechanism:

[0079] When the user's current emotional state is identified as positive emotions such as positive / happy / excited, the AI ​​outputs content that strengthens this emotional tendency.

[0080] For example: by making the language style more emotional, the plot more motivating, and the images brighter and more vivid, the user's existing creative motivation is amplified.

[0081] 2. Negative resonance mechanism:

[0082] When the user is identified as being in a negative emotion (such as anxiety, drowsiness, depression, anger) or lack of inspiration, the AI ​​actively generates suggestions with an "emotion regulation effect".

[0083] For example: the output content can be biased towards being light and humorous, encouraging and positive, with a slow pace and clear structure, to help the user's emotions return to normal or their inspiration be rekindled.

[0084] S104. Determine the content of creative suggestions based on the emotional style template.

[0085] In some embodiments, the emotional style template includes at least one of semantic features, intonation style, pragmatic rhythm, and sentence structure;

[0086] Determining the content of creative suggestions based on the emotional style template includes:

[0087] Generating content of creative suggestions from at least one dimension of language style, creative structure, and emotional resonance based on the emotional style template.

[0088] Based on the above steps S101 to S104, the human-computer collaborative creation method of this application embodiment can obtain the targetThe physiological characteristics of the target object are used to determine the target object's emotional state information. Then, based on the emotional state information, the resonance direction of the target object is determined. Based on the resonance direction, an emotional style template corresponding to the resonance direction is obtained from a preset suggestion generation strategy library. Finally, based on the emotional style template, the creative suggestion content is determined. Therefore, this application embodiment can incorporate the user's real cognitive state and emotional feedback into the creative suggestion content generation mechanism. Based on the user's emotions, the corresponding creative suggestion content can be given, improving the efficiency of human-computer collaboration and thus improving the user experience.

[0089] This application embodiment can select templates of different generation styles according to the suggestion generation strategy library. For example, positive resonance prioritizes the generation of "plot-rising", "victory-preset", and "romantic-inspiring" content; negative resonance prioritizes the generation of "rhythm buffer", "emotional care", and "new beginning introduction" content.

[0090] This application provides two examples:

[0091] Example 1, positive resonance loop: When the user is excited, the AI ​​suggests: "You have established a character conflict. In the next part, would you consider introducing an extreme choice to drive the turning point?"

[0092] Example 2, negative resonance loop: When the user is anxious / stagnant, the AI ​​suggests: "Should we slow down the pace and use a monologue by the character to re-examine the plot?"

[0093] In some embodiments, after determining the content of the creative suggestions based on the emotional style template, the method further includes:

[0094] Obtaining feedback information from the target object regarding the content of the creative suggestions; wherein the feedback information includes at least one of the following: response behavior information and reading status information. The response behavior information includes any one of adopting the content of the creative suggestions, skipping the content of the creative suggestions, or modifying the content of the creative suggestions. The reading status information includes at least one of the voice signal and facial expression features of the target object when reading the content of the creative suggestions;

[0095] Adjusting the call weight of the emotional style template based on the feedback information; wherein the call weight is used to represent the probability that the emotional style template is selected.

[0096] Based on the above technical solution, the embodiments of this application are not static recommendations or emotion classification outputs, but rather feed the emotion judgment results back to the AI ​​generation system in real time, forming a content adaptation and control logic with human psychological adjustment as the core objective, constituting a dynamic resonance closed loop of "recognition-generation-feedback-learning". The embodiments of this application can significantly improve the user stickiness and empathy of AI in the fields of emotional creation, psychological healing, and interactive design.

[0097] Referring to Figure 2, the embodiments of this application provide a flowchart of another human-computer collaborative creation method. As shown in Figure 2, the human-computer collaborative creation method of the embodiments of this application includes: steps S201 to S204.

[0098] S201. Obtain the physiological characteristic information of the target object, and determine the emotional state information of the target object based on the physiological characteristic information.

[0099] S202. Determine the resonance direction of the target object based on the emotional state information: wherein, the resonance direction includes positive resonance and negative resonance, positive resonance is used to indicate that the emotional state of the target object is positive, and negative resonance is used to indicate that the emotional state of the target object is negative.

[0100] S203. Based on the resonance direction, obtain the emotional style template corresponding to the resonance direction from the preset suggestion generation strategy library.

[0101] S204. Determine the creative suggestion content based on the emotional style template.

[0102] Wherein, steps S201 to S204 in this embodiment of the application are consistent with the principle of steps S101 to S104, and will not be repeated here.

[0103] S205. Obtain the feedback information of the target object on the creative suggestion content.

[0104] The feedback information includes at least one of the following: response behavior information and reading status information. The response behavior information includes any one of adopting the creative suggestion content, skipping the creative suggestion content, and modifying the creative suggestion content. The reading status information includes at least one of the voice signal and facial expression features of the target object when reading the creative suggestion content.

[0105] S206. Adjust the call weight of the emotion style template based on the feedback information.

[0106] The call weight is used to represent the probability that the emotion style template is selected.

[0107] In some embodiments, adjusting the call weight of the emotion style template based on the feedback information includes:

[0108] Inputting the emotion style template and feedback information into the resonance enhancement model to obtain the fit score; wherein, the resonance enhancement model is trained based on training samples, and the training samples include the mapping relationship between the emotion style template and the feedback information and the corresponding fit score;

[0109] Adjusting the call weight of the emotion style template based on the fit score.

[0110] In this embodiment, the resonant effect of the suggestion can be scored for its suitability. If the suitability score is high, the prediction is successful, and the generation strategy can be saved with weights. If the suitability score is low, the prediction fails, the error is updated, and the emotional style template matched with the suggestion is adjusted.

[0111] The resonance enhancement model in this embodiment is an "individualized emotion-content mapping model," which aims to enable AI not only to understand the current emotional state of the user, but also to gradually build a personalized "emotion-content resonance structure" based on the user's past preferences for suggestions under different emotions, so as to improve the fit of the generated suggestions and the sense of creative symbiosis.

[0112] The resonance enhancement model in this embodiment can be applied to the following model structure and training method: using a dual-tower structure.(SiameseEncoder) encodes the "emotional state vector" and "suggestion content vector" respectively; it is trained using contrastive loss or triplet loss, with the goal of maximizing the similarity between the emotional state and the suggestion content vector if the suggestion is adopted, and minimizing the similarity if it is rejected. In addition, facial and voice data collected during the reading stage will be used as auxiliary labels for "suggestion adoption attitude" to improve the stability and credibility of the training set feedback signal.

[0113] Optionally, each user in this application embodiment has an independent "emotion-content preference mapping table"; in future creation, when a certain emotional state is identified, the personalized model will be called to prioritize the generation of suggestions that conform to their past preferences; the model will be continuously adjusted according to the feedback to build a dynamically evolving individualized creative style response system.

[0114] The resonance enhancement model in this application embodiment realizes a complete closed loop from "user emotional state" → "suggested content generation" → "acceptance behavior" → "model self-optimization", constructing a collaborative mechanism between humans and machines that is truly based on long-term resonance preferences, and is the core emotional engine of the entire symbiotic creation method.

[0115] This application embodiment can update the suggestion generation strategy library after each round of creation, increase the priority of high-fit templates, and reduce the activation probability of low-feedback templates; and synchronously update the matching weight of emotional type and suggestion type in the resonance direction.

[0116] This application embodiment introduces a "personalized resonance tuning mechanism" (PRT) into the human-computer symbiotic creation system to solve the problem that different users have significant individual differences in emotional response, language style preference, content acceptance degree, etc. This mechanism continuously collects user feedback, establishes an individualized resonance enhancement model, and dynamically adjusts the suggestion generation logic to achieve higher psychological fit and emotional resonance depth.

[0117] The resonance enhancement model of this application embodiment has a resonance optimization mechanism based on the following three core principles:

[0118] 1. Differences in user emotional responses: Different users have different emotional resonance intensities with the same suggestion template (e.g., "expressing emotions from a supporting character's perspective");

[0119] 2. Time-series characteristics of resonance feedback: Some suggestions have a delayed or phased effectiveness for users;

[0120] 3. Interaction preferences are related to creative style: Users' preferences when adopting suggestions (positive / negative adoption, whether to modify, etc.) are highly correlated with their creative language style, rhythm selection, and color scheme preference.

[0121] Therefore, the resonance optimization mechanism of this application embodiment needs to model the feedback as a multi-dimensional behavior-emotion fusion signal, continuously learn it, and then act in reverse on the suggestion generation module.

[0122] As an example, the technical implementation process of this application embodiment performs the following operations:

[0123] (1) Initialization stage: Specification 7 / 18 pages 10 CN 121093034 A

[0124] Based on the multi-dimensional emotion state vector E_user_t output by the emotion recognition module, abbreviated as: E_t;

[0125] Establish a mapping pair (T_i,E_t) with the current suggestion template T_i, and record the initial output fit score R_0.

[0126] (2) Feedback collection stage:

[0127] Collect user adoption behavior (whether to click to use, whether to modify and reuse);

[0128] During the user's reading of the content, collect physiological response (heart rate fluctuation, brain frequency conversion), facial dynamics (facial muscle activity), and speech response (speech speed / pitch change);

[0129] Integrate into behavior-emotion feedback vector F_{i,t}.

[0130] (3) Modeling of the resonance enhancement model:

[0131] A lightweight regression model (such as XGBoost or LSTM) is used to establish a mapping function between the feedback vector and the suggestion type: R_{i,t}=f(T_i,E_t,F_{i,t})+ε; This function outputs the “individualized fit score” R_{i,t} of the suggestion template, which is used to adjust the calling probability of the emotional style template in real time.

[0132] As an example, this application embodiment provides a complete process flow of signal acquisition—emotion recognition—suggestion generation—resonance feedback, including the following four stages:

[0133] First stage: signal acquisition

[0134] The user's physiological and behavioral data are collected in real time through multimodal sensing devices, including but not limited to: electroencephalogram (EEG): used to identify attention state and potential for inspiration bursts; heart rate and heart rate variability (PPG): reflecting the user's current emotional tension and rhythm stability; skin conductance (GSR): detecting the intensity of emotional activation; facial expression (Video-based FACS): monitoring micro-expression changes and muscle dynamics; speech signal (Prosody & Pitch): identifying tone, speech rate, and pitch changes; user text input / sketching behavior: monitoring creative language style and structural complexity.

[0135] Wherein, all signals are time-stamped and sent to the fusion processing module for preprocessing (noise reduction, standardization, and fragmentation processing).

[0136] Second Stage: Emotion Recognition

[0137] The preprocessed data to be processed is fed into the emotion recognition model, and classification and vector construction are performed using Support Vector Machine (SVM) or Long Short-Term Memory Network (LSTM).

[0138] The output is a multi-dimensional emotion state vector, including the following dimensions: arousal, affinity, directionality, stability, and individual difference mapping coefficient.

[0139] The recognition results are not only used to determine the emotion type in real time, but also support the construction of time series of "emotion change trajectory".The model is used to determine whether the "creative stagnation zone" or the "inspiration burst zone" has been entered.

[0140] Third stage: Suggestion generation

[0141] The multi-dimensional emotional state vector will be sent to the resonance direction determination module and classified as: positive resonance: the system provides push suggestions to stimulate creativity; negative resonance: the system provides soothing suggestions to repair creative emotions; neutral resonance: provides structural or perspective expansion suggestions.

[0142] The suggestion generation module calls the template engine to generate creative suggestions such as language, image, video, and melody based on the emotion type and contextual semantics.

[0143] For example: If the user's current emotion is "calm - confiding", the system can give: "Try to tell this story again from the perspective of an unfamiliar character, and you may find another ending."

[0144] Fourth stage: Resonance feedback

[0145] After the user adopts the suggestion, the system continuously listens for the user's feedback signals, which are divided into two categories:

[0146] Explicit feedback: Whether the user adopts the suggestion (click / drag in draft / fine-tune content); Manual 8 / 18 pages 11 CN 121093034 A

[0147] Implicit feedback: The user's tone activity, facial expression relaxation, and EEG activity frequency changes when reading the generated content.

[0148] The feedback information is structured and stored and then sent to the resonance enhancement module to train the personalized resonance model, adjust the future template call weight, and continuously optimize the "psychological fit". Finally, a complete operating loop is formed, which supports continuous learning, personalized adaptation and high-frequency co-creation. This is the key difference between the embodiments of this application and the traditional prompt generation system.

[0149] Referring to Figure 3, this application embodiment provides a flowchart for obtaining an emotional style template corresponding to the resonance direction. As shown in Figure 3, based on the resonance direction, an emotional style template corresponding to the resonance direction is obtained from a preset suggestion generation strategy library, including steps S301 to S302.

[0150] S301: Obtain the current creative content and extract context tags from the current creative content; wherein, the context tags include at least one of scene keywords, emotional themes, and plot nodes.

[0151] S302: Based on the resonance direction and context tags, an emotional style template corresponding to the resonance direction is obtained from a preset suggestion generation strategy library.

[0152] The human-computer collaborative creation method provided in this application embodiment can be applied to terminal devices such as mobile phones, tablets, wearable devices, in-vehicle devices, augmented reality (AR) / virtual reality (VR) devices, laptops, ultra-mobile personal computers (UMPCs), netbooks, and personal digital assistants (PDAs). This application embodiment provides specific methods for terminal devices.The body type is not limited in any way.

[0153] It should be understood that the number of each step in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0154] The embodiments of this application adopt the technical idea of ​​resonance-suggestion-feedback structure. Although the core application scenario is designed with creative inspiration and psychological healing, the structure has high modularity and adaptability, and can also be extended to other scenarios based on user state recognition and personalized suggestion output, including but not limited to: (1) situational incentive mechanism in educational feedback system; (2) empathic task guidance in career planning assistance; (3) emotional synchronization and expression support in social companionship system.

[0155] It should be noted that the technical implementation path of the embodiments of this application does not change substantially. It only loads different content data sources and target task models according to different scenarios, which is a reasonable extension of the technical solution of this application.

[0156] The embodiments of this application can be applied to the following scenario types and recommended expression methods.

[0157]

[0158] This application embodiment is an AI proactive suggestion generation mechanism based on emotion recognition and creative context. It aims to automatically push context-adaptive creative suggestions to users when they experience states such as "low inspiration", "unstable emotions" or "creative stagnation" during the creative process, without requiring the user to actively request them. This restores the creative rhythm and enhances the continuity and emotional support of human-computer co-creation.

[0159] The core of the human-computer collaborative creation system in this application embodiment is: after detecting an inefficient creation state, it actively calls the suggestion generation module to generate suggestion content that integrates language style, creative perspective and emotional resonance, including the following content:

[0160] (1) Trigger judgment unit

[0161] Based on the input of the multimodal emotion recognition system, the inefficient state is judged by combining the following features:

[0162] EEG: low frequency amplitude increases;

[0163] PPG: heart rate variability (HRV) decreases;

[0164] Speech signal: tone tends to be stable and speech rate decreases;

[0165] Facial expression features: muscle tension decreases and facial expression fluctuations weaken.

[0166] When the above features trigger the preset threshold, the system sets the trigger flag T=1 and enters the suggestion generation process.

[0167] (2) Context tag extraction unit

[0168] Semantic analysis is performed on the current creative text, sketch or audio recording, and the following contextual tags are automatically extracted:

[0169] Scene keywords (such as "rainy night" "cabin in the deep forest");

[0170] Emotional themes (such as "loneliness", "depression" "nostalgia");

[0171] Plot nodes (such as "the night before separation", "emotional turning point", "intensification of conflict").

[0172] Contextual tags serve as prompt vectors in the emotional style template matching process.

[0173] (3) Suggestion content generation unit

[0174] Combining emotional state information and contextual tags, the most matching emotional style template is called from the suggestion generation strategy library, and the suggested content with features including but not limited to the following:

[0175] Language style dimension: poetic, calm narrative, humorous and slow-release, etc.;

[0176] Creative structure dimension: role perspective switching, non-linear time progression, environmental emotion reversal, etc.;

[0177] Emotional resonance dimension: motivational, comforting, guiding, etc.

[0178] As an example: Example output (context: creator is in a low mood, draft is emotionally empty):

[0179] "If you are facing a wall of powerlessness and silence right now, perhaps you can start from the voice behind the wall - who is calling from the other side? Let us start from that echo."

[0180] This suggestion resonates between language and context, stimulates the user's creative inspiration through style guidance, and completes an emotionally driven content restart.

[0181] All actively generated suggestions and their user feedback (such as whether they are adopted, reading responses, etc.) in this application embodiment will enter the strategy optimization module to train the resonance enhancement model, improve the matching accuracy of future suggestions in terms of emotional fit and content acceptance rate, and build a continuously learning co-creation system.

[0182] This application embodiment is a human-computer symbiotic creation method and system based on multimodal emotional feedback and suggestion generation, which can be widely applied to various creation types. This application embodiment provides five core application scenarios and their respective typical implementation paths and system adaptation strategies.

[0183] (1) Literary creation (novels, scripts, poems)

[0184] Key points of the acquisition module: EEG+GSR+text generation features

[0185] Key points of resonance recognition: monitoring "inspiration blockage" and "emotional outburst zone", such as a sudden decrease in paragraph length and an increase in the proportion of α / θ waves in EEG; Instruction manual 10 / 18 pages 13 CN 121093034 A

[0186] Suggestion generation type: context jump suggestion (e.g. "write it again on a different timeline"); writing style suggestion (e.g. "try to rewrite the previous paragraph using Hemingway's minimalist language");

[0187] Feedback acquisition method: text adoption rate, modification range, and the activity level of voice and facial expressions when reading the content.

[0188] System output: Personalized writing style template update, long-term training to generate style adaptation curves (e.g., "subjective immersive" vs. "observer structure" creators);

[0189] (2) Visual design (illustration, game characters, advertising sketches)

[0190] Key points of the acquisition module: facial expressions + EEG + cursor / touch path;

[0191] Key points of resonance recognition: repeated undoing behavior during the drawing process; facial expression suppression + GSR decrease.

[0192] Suggested generation types: Tone reconstruction suggestions (e.g., "Try warm and cool contrast"); Composition center of gravity prompts (e.g., "Enlarge the character by 10% and center it");

[0193] Feedback collection methods: Whether to directly use the sketch generated by the suggestions; Drawing speed and facial changes (e.g., slight frown - slight relaxation);

[0194] System output: Composition preference model; Color tone emotion preference model (e.g., neutral color vs. saturated contrasting color);

[0195] (3) Video creativity (short video script, storyboard sketch)

[0196] Collection module focus: Voice + expression + sketch + semantic paragraph rhythm analysis;

[0197] Resonance recognition key points: Storyboard sketch modification frequency + voice tone continuously decreasing;

[0198] Suggested generation types: Shot rhythm suggestions (e.g., "Add a one-second freeze delay to enhance emotion"); Transition connection suggestions (e.g., "Switch perspective narrative from dream"); ​​

[0199] Feedback collection method: the ratio of suggested frames adopted by the user on the timeline; the change in speech rate when reading the script; system output: rhythm-emotional resonance mapping emotional style template; style classification (editing type / atmosphere type / drama type) automatic suggestion preference;

[0200] (4) Music composition (lyrics, melody suggestions)

[0201] Collection module focus: EEG + speech + rhythm synchronization;

[0202] Resonance recognition key points: heart rate and EEG synchronous resonance state detection (inspiration zone);

[0203] Suggestion generation type: next melody suggestion (e.g. "key rise into chorus"), style conversion suggestion (e.g. "light jazz to lyrical electronic");

[0204] Feedback collection method: whether the user adopts the melody suggestion and plays it; the change in speech energy spectrum density when singing;

[0205] System output: music style-emotional template matrix; user preference curve is used to enhance the accuracy of melody recommendation;

[0206] (5) Treatment and psychological counseling (trauma expression, empathy intervention)

[0207] Key points of the acquisition module: GSR + heart rate + facial muscle group activity (FACS);

[0208] Key points of resonance recognition: Recognizing emotional suppression, sadness, and emotional avoidance patterns;

[0209] Suggestion generation type: Empathic question (e.g., "Do you want to write a sentence to your past self?"); Trauma image transformation suggestion (e.g., "Concretize the pain as a tree");

[0210] Feedback acquisition method: Physiological changes before and after user expression (heart rate fluctuation amplitude); frequency of smiling and changes in eye muscle tension after expression;

[0211] System output: Step-by-step guided empathy template; Individualized trauma expression path modeling (non-linear release vs. layer-by-layer backtracking).

[0212] This application embodiment can also continuously monitor the creator's multimodal emotional signals and language generation behavior in real time during the human-computer co-creation process. When the following feature combination is detected, the system will enter the "suggestion intervention" state:

[0213] Abnormal physiological signals: The proportion of α / θ in the frequency band of EEG signals is abnormally increased (manifested as decreased attention), heart rate fluctuation tends to stabilize (low stress), and GSR conductance activity is significantly reduced;

[0214] Behavioral feedback signals: The reading speed is reduced, the tone is flattened, and the facial expression activity is reduced;

[0215] Text / draft analysis: The length of the generated content is shortened in a short time, the emotional color is neutral, and the syntactic structure tends to be mechanical.

[0216] Based on the judgment of the above multimodal fusion model, the active suggestion mechanism is triggered. The specific process is as follows:

[0217] (1) The semantic recognition model (such as BERT + sentiment classification network) analyzes the scene intention of the creator's current input text as "character isolation + conflict not erupted";

[0218] (2) It is determined that the resonance direction is "reverse resonance need" (i.e., need for comforting intervention);

[0219] (3) The system selects a template with the emotional style "gentle tone" from the suggestion generation strategy library and calls the prompt generation model (PromptedSuggestion Module) to generate the following suggestion:

[0220] "When the character chooses to remain silent, let the wind speak for her. Write a letter that she never sent to the person she once wanted to leave."

[0221] This suggestion has the following characteristics: Emotional style: soft tone, indirect expression, implied emotional transition; Content strategy: Using "letters" as a psychological externalization tool, the character is guided to express emotions; guidance mode: providing the direction of the character's behavior, while opening up specific content details, stimulating the creator to reconstruct the narrative momentum.

[0222] After the user receives the suggestion, they can continue to create based on this semantic path, and their feedback data (including micro-expressions and voice changes during the reading process) will be included in the next round of weight optimization to form positive training data for personalized recommendations.

[0223] This application embodiment can be extended to various creative scenarios such as script creation, novel writing and psychological writing training.

[0224] Referring to Figure 4, this application embodiment provides a structural schematic diagram of a human-computer collaborative creation system. As shown in Figure 4, the human-computer collaborative creation system of this application embodiment includes a signal acquisition layer 41, a preprocessing and fusion node 42, a main control computing module 43 and a user interface module 44.

[0225] The signal acquisition layer 41 includes an EEG collector 411, a heart rate sensor 412, a skin conductance probe 413, a camera 414 and a microphone 415.

[0226] Specifically, the EEG acquisition device 411 acquires an electroencephalogram (EEG), the heart rate sensor 412 acquires a heart rate pulse (PPG), the skin conductance probe 413 acquires a skin surface rhythm (GSR), the camera 414 acquires facial expressions, and the microphone 415 acquires speech signals.

[0227] Optionally, the human-computer collaborative creation system of this application embodiment has a Bluetooth or Wi-Fi wireless transmission interface, supporting...Low-power continuous data acquisition, with all acquisition devices supporting a unified driver protocol and timestamp synchronization.

[0228] Optionally, the preprocessing and fusion node 42 can be equipped with an edge processor (such as NVIDIA Jetson Nano or Raspberry Pi 5) to perform preliminary data cleaning, feature extraction, and standardization processing; at the same time, the preprocessing and fusion node 42 has a built-in small data caching mechanism to ensure data integrity in the event of a network outage; the preprocessing and fusion node 42 supports low-latency feature fusion.

[0229] Optionally, the main control computing module 43 is equipped with a host server or a local high-computing device, integrating an emotion recognition model (SVM, LSTM, Transformer, etc.), a resonance determiner, and a suggestion generation engine, which can achieve cross-platform compatibility by deploying the model.

[0230] Optionally, the user interface module 44 includes a terminal display, a touch interface, or a mobile app, used to display suggested content, created content, collect user feedback (clicks, expressions, voice, etc.) and send it to the core module to optimize model weights.

[0231] The human-computer collaborative creation system of this application embodiment is designed with open standard interfaces. To ensure system scalability and module interoperability, the human-computer collaborative creation system of this application embodiment adopts the following standardized API interface set:

[0232] EEG_API.init(device_id): Initializes the EEG device; Specification 12 / 18 pages 15 CN 121093034 A

[0233] HR_API.get_baseline(): Gets the baseline fluctuation of heart rate;

[0234] EMO_API.predict(input_vector): Returns the current emotional state vector;

[0235] SUG_API.generate(scene_tag,emo_state): Generates suggested content based on context and emotion;

[0236] FDB_API.collect(user_id,expression_data,audio_data): Collects feedback and tags it;

[0237] ADAPT_API.update_weights(uid,feedback_score): Updates weights according to feedback;

[0238] The API interface of this application uses a RESTful + WebSocket hybrid protocol, supports local deployment and cloud calls, and reserves multi-language SDKs (Python / JavaScript / Swift) for third-party integration.

[0239] The human-computer collaborative creation system of this application mainly demonstrates the modular signal acquisition and calculation system architecture and standardized API interface design, and proposes to adopt a modular structure at the hardware level to achieve high scalability and adaptability to multiple systems.Diversified creation scenarios and algorithm model integration requirements. The human-computer collaborative creation system of this application embodiment has plug-and-play, hot-swap recognition, edge computing and remote communication capabilities.

[0240] The human-computer collaborative creation system of this application embodiment is a human-computer symbiotic creation system, which designs a "dynamic dominance allocation mechanism" to dynamically adjust the participation and dominance of AI and humans in the creation process according to the creator's real-time emotional state, cognitive load and interactive behavior characteristics, thereby optimizing creation efficiency and experience.

[0241] In order to further improve the performance of the human-computer collaborative creation system of this application embodiment in terms of emotion recognition accuracy and response sensitivity, the human-computer collaborative creation system of this application embodiment supports compatibility and adaptation with brain-computer interface (BCI) devices.

[0242] BCI technology can achieve more refined modeling of psychological variables such as user cognitive state, emotional fluctuations and inspiration bursts through high-precision electroencephalogram (EEG) acquisition and processing. Typical devices include, but are not limited to, non-invasive BCI systems that support multi-channel high-density sampling (such as OpenBCI, NextMind, Emotiv, etc.).

[0243] In the signal acquisition layer 41 of the human-computer collaborative creation system of this application embodiment, a standard interface for accessing BCI devices (such as BLE, USB, SDK / API protocol layer) is reserved, which can realize the following functions: (1) dynamic monitoring and time synchronization of EEG frequency bands (α / β / θ / γ); (2) joint modeling of inspiration activation index and emotional stability; (3) coupling analysis of brain region activity map and suggestion adoption behavior; (4) real-time resonance feedback mechanism under high sampling rate.

[0244] The core innovation of the human-computer collaborative creation system of this application embodiment lies in the suggestion generation mechanism and personalized resonance optimization algorithm constructed based on high-precision emotional state parameters. As one of the optional acquisition entry points of the human-computer collaborative creation system of this application embodiment, the introduction of the brain-computer interface device can further expand the applicability and industrial depth of the system in high-demand scenarios such as clinical psychology, neuro-art, and trauma intervention.

[0245] In the user experience (UX) and interface interaction (UI) design of the human-computer collaborative creation system, this application embodiment constructs a highly human-like five-stage interaction path, enabling users to collaborate with the system naturally and smoothly even without a programming background.

[0246] 1. Five-stage user interaction process

[0247] (1) Creation initiation: After entering the creation interface, the user selects the creation type (such as novel, lyrics, illustration, etc.) and inputs the initial content (text, sketch, melody, etc.). The system simultaneously starts the multimodal signal acquisition module (EEG, GSR, facial expression, voice, etc.).

[0248] (2) Emotion mapping: The "emotional state dashboard" is constructed and visualized in real time, and the feedback includes: current focus curve,Heart rate fluctuations, emotional trend graphs, and identified resonance directions (such as "motivational expectation" or "calm reassurance").

[0249] (3) Suggestion reception: When the system determines that the user has entered a creative bottleneck or emotional fluctuation range, a "resonance suggestion card" will automatically pop up, which includes: scene prompts (such as "reconstruction from the perspective of the opposing character"); style suggestions (such as "calm narration / introspective style instruction manual 13 / 18 pages 16 CN 121093034 A style"); multimodal content fragments (such as structural sketches, melody fragments, reference images, etc.). The user can choose "adopt", "skip" or "adopt after modification".

[0250] (4) Feedback collection: User feedback is collected in various ways, including:

[0251] explicit behavior: click behavior, dragging content to the draft area;

[0252] implicit signals: tone activity during reading, facial micro-expressions, EEG frequency increase, etc. The feedback will be used to adjust the suggestion template weight and resonance model parameters in real time.

[0253] (5) Resonance enhancement model and content fusion: Feedback data is used to train personalized resonance adaptation curves and dynamically called and optimized in the next round of suggestion generation. Some high-scoring suggestions will be included in the public corpus and used for model retraining with user authorization to enhance the overall generation quality of the platform.

[0254] The human-computer collaborative creation system of this application embodiment can also be embedded into commonly used creation tools (such as Photoshop, Notion, Obsidian). This application embodiment can realize: low interference mode, only presenting the emotional trajectory map without pop-up interruption; silent observation mode, where the AI ​​does not output suggestions, but only passively observes, for scientific research control experiments or psychological analysis.

[0255] The human-computer collaborative creation system of this application embodiment is an application embodiment that uses the resonance enhancement module to dynamically adjust the language style, image tone and rhythm parameters to achieve high "psychological adaptation". The "resonance enhancement module" proposed in this application embodiment is located between suggestion generation and user feedback, as an emotional refinement and optimization mechanism in the human-computer co-creation process. Its core function is: based on the user's emotion recognition results and feedback signals, dynamically adjust the expression style and sensory form of the AI ​​output content, improve the adaptability and resonance between the creative suggestions and the user's psychological state, thereby enhancing the acceptance and emotional penetration of the system-generated content.

[0256] As an example, the embodiments of this application further illustrate the following four aspects:

[0257] I. System Mechanism Overview: by continuously collecting the following signals from the user during the writing or reading process:

[0258] Micro-expression dynamics (such as eyelid tension, mouth corner movement);

[0259] Speech signals (speech rate, tone change, fundamental frequency distribution);

[0260] Physiological indicators (EEG frequency band distribution, heart rate fluctuation, skin conductance);

[0261] Combining the contextual semantic information of the suggestion content, the system evaluates the "heart" relationship between the current suggestion and the user's emotional state."Reasonable fit", and trigger resonance enhancement process when necessary: ​​

[0262] II. Typical application scenario: music lyrics creation background;

[0263] When a user tries to create lyrics for a melody, the system detects the following states:

[0264] Facial signals: furrowed brows, still expression → indicating sadness;

[0265] Voice signals: low fundamental frequency, flat tone, slow speech rate;

[0266] Physiological signals: EEG dominated by α / θ frequency, small heart rate fluctuation, low GSR activity → overall low arousal depression state.

[0267] The system comprehensively judges that the user is in the "sadness-reflection" type of resonance direction, and automatically starts the following multimodal resonance enhancement measures.

[0268] III. Resonance enhancement operation:

[0269] 1. Fine adjustment of language style:

[0270] Replace the external narrative sentences in the original suggestion with introspective monologue expressions;

[0271] Example: Original sentence: "He walked in the rain, without looking back." "

[0272] Optimization: "I was still standing there, listening to his footsteps walking away." "

[0273] 2. Image tone adjustment (for illustrations or MV sketches):

[0274] Color style: transition from bright blue-green to cool gray with low saturation; Instruction manual 14 / 18 pages 17 CN 121093034 A

[0275] Scene atmosphere: change from "open street" to "nighttime car window reflection".

[0276] 3. Rhythm and beat adjustment:

[0277] Recommend a lyrical structure in 2 / 4 time;

[0278] Insert a "weak start" type rhythm template to enhance the emotional fit between the sentences and the melody.

[0279] 4. Psychological fit feedback verification:

[0280] After the user reads the suggested content, the system detects an increase in the activity of their facial expressions, an increase in speech speed, and a shift of the brainwave frequency band to the mid-frequency;

[0281] It is determined that the suggestion resonates highly with the emotions;

[0282] This round of data is used as a "positive sample" feedback model to train the system for personalized resonance strategy updates.

[0283] IV. Application Scenarios Expansion:

[0284] This application is applicable to the following multimodal creation types:

[0285] Literary writing: Optimizing the consistency between paragraph tone and character emotions;

[0286] Illustration design: Adjusting character dynamics and background style to match narrative emotions;

[0287] Video script: Rewriting dialogue rhythm and scene sequence to match emotional tension;

[0288] Music creation: Adjusting lyric rhythm and melody resonance to achieve high emotional connectivity.

[0289] V. Summary of Technical Points:

[0290] The resonance enhancement module in this application embodiment achieves: output optimization with user emotions as a dynamic control factor; establishing a high-frequency resonance adjustment closed loop from feedback signal → output content → emotional change → model update; making AI suggestions closer to the human perception system in the emotional dimension, and promoting human-machine co-creation towards deep emotional collaboration.

[0291] Referring to Figure 5, this application embodiment provides a structural schematic diagram of a human-computer collaborative creation device 50. As shown in Figure 5, the human-computer collaborative creation device 50 of this application embodiment includes: a first determining module 501, a second determining module 502, a template acquisition module 503, and a third determining module 504.

[0292] The first determining module 501 is used to acquire the physiological characteristic information of the target object, and determine the emotional state information of the target object based on the physiological characteristic information;

[0293] The second determining module 502 is used to determine the resonance direction of the target object based on the emotional state information: wherein, the resonance direction includes positive resonance and negative resonance, positive resonance is used to indicate that the emotional state of the target object is positive, and negative resonance is used to indicate that the emotional state of the target object is negative;

[0294] The template acquisition module 503 is used to acquire the emotional style template corresponding to the resonance direction from a preset suggestion generation strategy library based on the resonance direction;

[0295] The third determining module 504 is used to determine the content of the creation suggestion based on the emotional style template.

[0296] Optionally, the human-computer collaborative creation device 50 in this application embodiment further includes an adjustment module, which is used to obtain feedback information from the target object regarding the creation suggestion content; wherein, the feedback information includes at least one of the following: response behavior information and reading status information, wherein the response behavior information includes any one of adopting the creation suggestion content, skipping the creation suggestion content, and modifying the creation suggestion content, and the reading status information includes at least one of the voice signal and facial expression features of the target object when reading the creation suggestion content; and the call weight of the emotional style template is adjusted based on the feedback information; wherein, the call weight is used to represent the probability of the emotional style template being selected.

[0297] Optionally, the adjustment module is also used to input the emotional style template and feedback information into the resonance enhancement model to obtain a fit score; wherein, the resonance enhancement model is trained based on training samples, and the training samples include the mapping relationship between the emotional style template and the feedback information and the corresponding fit score; and the call weight of the emotional style template is adjusted based on the fit score.

[0298] Optionally, the first determining module 501 is specifically used to input physiological feature information into the fusion processing module for signal preprocessing to obtain data to be processed; wherein, the fusion processing module's fusion method includes at least one of feature-level splicing fusion, attention-based weighted fusion, and gating-based dynamic fusion; inputting the data to be processed into the emotion recognition model to obtain a multi-dimensional emotion state vector as emotion state information; wherein, the emotion recognition model is obtained based on a support vector machine, a long short-term memory network, or an ensemble classification model.

[0299] Optionally, the second determining module 502 is used to determine if the multi-dimensional emotion state vector belongs to a preset positive resonance interval,The resonance direction of the target object is determined to be positive resonance; if the multidimensional emotional state vector belongs to the preset negative resonance interval, the resonance direction of the target object is determined to be negative resonance.

[0300] Optionally, the template acquisition module 503 is specifically used to acquire the current creative content and extract the context tag from the current creative content; wherein, the context tag includes at least one of scene keywords, emotional theme and plot node; based on the resonance direction and context tag, the emotional style template corresponding to the resonance direction is acquired from the preset suggestion generation strategy library.

[0301] Optionally, the third determination module 504 is specifically used to generate creative suggestion content based on the emotional style template from at least one of the language style dimension, creative structure dimension and emotional resonance dimension.

[0302] In application, each module in the human-computer collaborative creation device 50 of this application embodiment can be a software program module, or it can be implemented by different logic circuits integrated in the processor, or it can be implemented by multiple distributed processors.

[0303] The human-computer collaborative creation device 50 of this application embodiment can execute the method provided in this application embodiment. The implementation principle is similar. The actions performed by each module in the device of each embodiment of this application correspond to the steps in the method of each embodiment of this application. For detailed functional descriptions of each module of the device, please refer to the descriptions in the corresponding methods shown above. They will not be repeated here.

[0304] Referring to FIG6, this application embodiment provides a schematic diagram of the structure of a terminal device 6. As shown in FIG6, the terminal device 6 of this embodiment includes: a memory 61, a processor 60, and a computer program 62 stored in the memory 61 and executable on the processor 60. When the processor 60 executes the computer program, it implements the steps of the method of each embodiment of this application.

[0305] The terminal device 6 can be a desktop computer, a laptop, a handheld computer, or a cloud server, etc. The terminal device 6 may include, but is not limited to, a processor 60 and a memory 61. Those skilled in the art will understand that Figure 6 is merely an example of terminal device 6 and does not constitute a limitation on terminal device 6. It may include more or fewer components than shown, or combine certain components, or different components, such as input / output devices, network access devices, etc.

[0306] The processor 60 may be a central processing unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), or field-programmable gate arrays (FPGAs).Array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor, etc.

[0307] In some embodiments, memory 61 can be an internal storage unit, such as a hard disk or memory. Memory 61 can be a removable / non-removable, volatile / non-volatile computer system storage medium, for example: memory 61 is a non-volatile memory used to read and write non-volatile magnetic media. In other embodiments, memory 61 can also be an external storage device, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc. equipped on terminal device 6. Memory 61 is used to store operating system, application program, bootloader, data and other programs, such as program code of computer programs, etc. Memory 61 can also be used to temporarily store data that has been output or will be output.

[0308] It should be noted that the information interaction and execution process between the above-mentioned devices / units are based on the same concept as the method embodiments of this application. Their specific functions and technical effects can be found in the method embodiments section, and will not be repeated here.

[0309] Those skilled in the art will understand that, for the sake of convenience and brevity, the above-mentioned division of functional units and modules is used as an example. In practical applications, the above-mentioned functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of each functional unit and module are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units and modules in the above system can be referred to the corresponding process in the aforementioned method embodiments, and will not be repeated here.

[0310] This application embodiment also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps in the above-described method embodiments.

[0311] If the above-described integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, this application implements the above-described embodiments...All or part of the processes in the method can be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium. When executed by a processor, the computer program can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable file, or some intermediate form. The computer-readable medium can include at least: any entity or device capable of carrying the computer program code to a device / terminal device, a recording medium, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, such as a USB flash drive, a portable hard drive, a magnetic disk, or an optical disk.

[0312] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk drive (HDD), or solid-state drive (SSD), etc. The storage medium can also include combinations of the above types of memory.

[0313] This application provides a computer program product. When the computer program product is run on a processor, the processor executes the steps in the above method embodiments.

[0314] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0315] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0316] In the embodiments provided in this application, it should be understood that the disclosed apparatus / network devices and methods can be found in the specification on pages 17 / 18 of 20 CN 121093034 A.It can be achieved in other ways. For example, the device / network device embodiments described above are merely illustrative. For example, the division of the modules or units described above is merely a logical functional division. In actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed may be through some interfaces, and the indirect coupling or communication connection of devices or units may be electrical, mechanical or other forms.

[0317] The units described above as separate components may or may not be physically separated. The components shown as units may or may not be physical units, that is, they may be located in one place or may be distributed on multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0318] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit it; although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent substitutions for some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application. Instruction manual, page 18 / 18, 21 CN 121093034 A, Figure 1; Instruction manual, Figure 1 / 4, page 22 CN 121093034 A, Figure 2; Instruction manual, Figure 2 / 4, page 23 CN 121093034 A, Figure 3; Figure 4; Instruction manual, Figure 3 / 4, page 24 CN 121093034 A, Figure 5; Figure 6; Instruction manual, Figure 4 / 4, page 25 CN 121093034 A HUMAN-COMPUTER COLLABORATIVE CREATION METHOD, APPARATUS, TERMINAL DEVICE, AND STORAGE MEDIUM ABSTRACT The present application provides a human-computer collaborative creation method, apparatus, terminal device, and storage medium, applicable to the field of human-computer interaction technology. The methodincludes: acquiring physiological characteristic information of a target object, and based on the physiological characteristic information, determining emotional state information of the target object; based on the emotional state information, determining a resonance direction of the target object; wherein the resonance direction includes positive resonance and negative resonance, positive resonance is used to indicate that the emotional state of the target object is a positive emotion, and negative resonance is used to indicate that the emotional state of the target object is a negative emotion; based on the resonance direction, from a preset suggestion generation strategy library, acquiring an emotional style template corresponding to the resonance direction; based on the emotional style template, determining creation suggestion content. The embodiments of the present application can provide corresponding creation suggestion content based on the user's emotion, and provide theefficiency of human-computer collaboration.

Claims

1. A human-computer collaborative creation method, characterized in that, include: Obtain physiological characteristic information of the target object, and determine the emotional state information of the target object based on the physiological characteristic information; Based on the emotional state information, the resonance direction of the target object is determined: wherein the resonance direction includes positive resonance and negative resonance, the positive resonance is used to indicate that the emotional state of the target object is positive, and the negative resonance is used to indicate that the emotional state of the target object is negative. Based on the resonance direction, obtain the emotional style template corresponding to the resonance direction from the preset suggestion generation strategy library; Based on the aforementioned mood style template, determine the suggested content for creation.

2. The human-computer collaborative creation method according to claim 1, characterized in that, After determining the suggested creative content based on the emotional style template, the process also includes: Obtain feedback information from the target object regarding the creative suggestion content; wherein the feedback information includes at least one of the following: response behavior information and reading status information, the response behavior information includes any one of adopting the creative suggestion content, skipping the creative suggestion content, and modifying the creative suggestion content, and the reading status information includes at least one of the voice signal and facial expression features of the target object when reading the creative suggestion content; Based on the feedback information, the call weight of the emotion style template is adjusted; wherein, the call weight is used to represent the probability of the emotion style template being selected and obtained.

3. The human-computer collaborative creation method according to claim 2, characterized in that, The step of adjusting the call weight of the mood style template based on the feedback information includes: The emotional style template and the feedback information are input into the resonance enhancement model to obtain a fit score; wherein, the resonance enhancement model is trained based on training samples, and the training samples include the mapping relationship between the emotional style template and the feedback information and the corresponding fit score; Based on the fit score, the call weight of the emotional style template is adjusted.

4. The human-computer collaborative creation method according to claim 1, characterized in that, The physiological characteristic information includes at least one of the physiological signals and behavioral signals of the target object collected during the human-computer collaborative creation process. The physiological signals include at least one of electroencephalogram (EEG), heart rate, and skin conductance response. The behavioral signals include at least one of the speech signals and facial expression features of the target object when reading the current creative content. Determining the emotional state information of the target object based on the physiological characteristic information includes: The physiological feature information is input into the fusion processing module for signal preprocessing to obtain the data to be processed; wherein, the fusion processing module includes at least one of feature-level splicing fusion, attention-based weighted fusion, and gating-based dynamic fusion. The data to be processed is input into the emotion recognition model to obtain a multi-dimensional emotion state vector, which is used as the emotion state information; wherein, the emotion recognition model is based on support vector machine or long short-term memory network or ensemble classification model.

5. The human-computer collaborative creation method according to claim 4, characterized in that, Determining the resonance direction of the target object based on the emotional state information includes: If the multidimensional emotional state vector belongs to the preset positive resonance interval, then the resonance direction of the target object is determined to be positive resonance; If the multidimensional emotional state vector belongs to the preset negative resonance interval, then the resonance direction of the target object is determined to be negative resonance.

6. The human-computer collaborative creation method according to any one of claims 1-5, characterized in that, The step of obtaining an emotional style template corresponding to the resonance direction from a preset suggestion generation strategy library based on the resonance direction includes: Obtain the current creative content and extract context tags from the current creative content; wherein, the context tags include at least one of scene keywords, emotional themes, and plot nodes; Based on the resonance direction and the context label, obtain the emotional style template corresponding to the resonance direction from the preset suggestion generation strategy library.

7. The human-computer collaborative creation method according to any one of claims 1-5, characterized in that, The emotional style template includes at least one of semantic features, intonation style, pragmatic rhythm, and sentence structure; The process of determining creative suggestions based on the emotional style template includes: Based on the aforementioned emotional style template, the creative suggestion content is generated from at least one of the dimensions of language style, creative structure, and emotional resonance.

8. A human-computer collaborative creation device, characterized in that, include: The first determining module is used to acquire physiological characteristic information of the target object and determine the emotional state information of the target object based on the physiological characteristic information. The second determining module is used to determine the resonance direction of the target object based on the emotional state information: wherein the resonance direction includes positive resonance and negative resonance, the positive resonance is used to indicate that the emotional state of the target object is positive, and the negative resonance is used to indicate that the emotional state of the target object is negative. The template acquisition module is used to acquire an emotional style template corresponding to the resonance direction from a preset suggestion generation strategy library based on the resonance direction. The third determining module is used to determine the content of the creative suggestions based on the emotional style template.

9. A terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1 to 7.

Images