Active AI digital human interaction method and system based on visual analysis

Abstract

The invention provides an active AI digital human interaction method and system based on visual analysis, and belongs to the technical field of artificial intelligence, and the method comprises the steps: 1, carrying out the batch extraction of image materials of a preset data source, carrying out the feature extraction of each batch of image materials, determining training detection parameters based on the extracted features, and determining a detection model; 2, collecting a picture of the target area in real time based on preset collection equipment, analyzing the picture in combination with the detection model, and determining whether to execute an interaction process or not based on a picture analysis result; 3, executing an interaction process, obtaining information consulted by the user for analysis, and meanwhile, collecting user actions in real time based on preset collection equipment and a detection model; 4, analyzing the user action, and further generating an answering scheme in combination with an analysis result; and step 5, obtaining a real-time picture analysis result, and further determining whether to execute a farewell process or not. Interaction intellectualization, real-time performance and user experience are improved.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to an active AI digital human interaction method and system based on visual analysis. Background Art

[0002] With the rapid development of artificial intelligence technology, AI digital humans, as an important carrier of human-computer interaction, have been widely used in education, medical care, entertainment, customer service and other fields. AI digital humans can provide users with an immersive interactive experience through multimodal interactions such as voice, vision and motion.

[0003] However, in the existing technology, the interactive ability of AI digital humans and users still has obvious limitations: the current AI digital humans mostly rely on active triggering by users, such as starting the interactive process through voice commands or key commands, and cannot achieve real-time environmental perception and active response, which greatly limits the naturalness and intelligence of the user experience. The existing AI digital humans have low accuracy when analyzing the emotional characteristics of users such as expressions, movements and tone. Especially in the fusion analysis of multimodal information, there is a lack of efficient algorithm support, resulting in an inaccurate understanding of user intentions, and the interactive process appears mechanical and stereotyped. In complex scenarios, AI digital humans cannot dynamically adjust the interactive strategy according to user behavior or context. For example, in the face of user emotional fluctuations or environmental changes, the response of existing systems is often delayed or mismatched, which affects the interactive effect and user satisfaction.

[0004] Therefore, there is an urgent need for an active AI digital human interaction method based on visual analysis to achieve accurate perception and real-time response to user emotions, actions and environment, improve the initiative, adaptability and interaction effect of AI digital humans, and meet users' interaction needs in diverse scenarios. Summary of the invention

[0005] The present invention provides an active AI digital human interaction method and system based on visual analysis, which are used to realize active AI digital human interaction through visual analysis combined with a detection model, solving the problems of passive interaction mode, delayed response and single-dimensional analysis in the prior art, and significantly improving the intelligence, real-time nature and user experience of the interaction by accurately extracting image features, real-time picture and user action analysis, and dynamically judging the interaction and farewell process, thereby ensuring the smoothness and integrity of the interaction process.

[0006] The present invention provides an active AI digital human interaction method based on visual analysis, comprising:

[0007] Step 1: Batch extract image materials from a preset data source, perform feature extraction on each batch of image materials, determine training detection parameters based on the extracted features, and then determine the detection model;

[0008] Step 2: Based on the preset acquisition equipment, the images of the target area are collected in real time, and the images are analyzed in combination with the detection model. Based on the image analysis results, it is determined whether to execute the interactive process;

[0009] Step 3: Execute the interactive process and obtain the user's consultation information for analysis. At the same time, collect user actions in real time based on the preset collection equipment and detection model;

[0010] Step 4: Analyze the user's actions and generate a solution based on the analysis results;

[0011] Step 5: Obtain the real-time image analysis results, and then determine whether to execute the farewell process.

[0012] Preferably, batch extraction is performed on image materials of a preset data source, feature extraction is performed on image materials of each batch, and training detection parameters are determined based on the extracted features, thereby determining a detection model, including:

[0013] Batch extracting image materials from a preset data source, classifying and labeling each batch of image materials, and then determining several groups of image materials;

[0014] Extracting a plurality of preset types of features from each group of image materials based on a first preset algorithm, and then determining a plurality of first feature sets;

[0015] Expanding each first feature set based on a second preset algorithm, thereby determining a plurality of second feature sets;

[0016] Extracting a plurality of third features from each second feature set based on a preset learning algorithm, and then determining the third features corresponding to each second feature set as a third feature set;

[0017] All third feature sets are used as training detection parameters, and a detection model is trained based on the training detection parameters, thereby determining the detection model.

[0018] Preferably, batch extraction of image materials from a preset data source includes:

[0019] Evaluate the image material of each preset data source to determine the stratification coefficient of each image material of each preset data source;

[0020] Determine the data level corresponding to each image material of each preset data source based on the hierarchical coefficient of each image material of each preset data source and the preset coefficient-level data table, and then divide the image material of each preset data source into a plurality of data levels;

[0021] Determine a number of sampling windows based on each data level of each preset data source and the quantum random number generator, and determine the scale of each sampling window;

[0022] Determine the size of the acquisition unit of each sampling window of each data level of each preset data source based on the scale of the sampling window of each data level of each preset data source and a preset scale-unit database;

[0023] The image materials of each data level of each preset data source are batch extracted based on the size of the acquisition unit of each sampling window of each data level of each preset data source.

[0024] Preferably, evaluating the image material of each preset data source to determine the stratification coefficient of each image material of each preset data source includes: ;

[0025] in, is the layering coefficient of the i-th image material of the u-th preset data source, is the total number of image features of the i-th image material of the u-th preset data source, is the g-th eigenvalue of the i-th image material of the u-th preset data source, is the feature mean of the i-th image material of the u-th preset data source, is the weight of the g-th feature of the i-th image material of the u-th preset data source, is the dynamic adjustment factor of the g-th feature of the i-th image material of the u-th preset data source, is the preset exponential magnification factor, is the preset smoothing factor, is the preset regularization factor.

[0026] Preferably, the image of the target area is collected in real time based on a preset collection device, the image is analyzed in combination with a detection model, and whether to execute the interactive process is determined based on the image analysis result, including:

[0027] The target area is locked based on the dynamic area detection algorithm, and the image of the target area is captured in real time based on the preset acquisition equipment;

[0028] Pre-process the captured images, and then analyze the captured images frame by frame based on the detection model to obtain the collected user information and determine whether the user actively sends an interactive signal;

[0029] Processing the collected user information to determine the amount of user information;

[0030] If it is detected that the user actively sends an interactive signal or the amount of user information is greater than or equal to the preset information threshold, the interactive process will be entered. If no interactive signal is detected or the amount of user information is less than the preset information threshold, the screen will continue to be monitored.

[0031] Preferably, processing the collected user information to determine the amount of user information includes:

[0032] Extract parameters from the collected user information, and then obtain relevant parameters of the user information volume;

[0033] The amount of user information is determined based on parameters related to the amount of user information, including: ;in, is the amount of user information, N is the number of users in the screen, is the number of time frames for action collection, For the The key points of the user's body in the frame, For the The key points of the user's body in the frame, is the number of user expression features in the picture, is the intensity value of the jth user’s facial expression feature, is the preset user motion amplitude conversion coefficient, It is the preset user expression intensity conversion coefficient.

[0034] Preferably, the user action is analyzed, and then a solution is generated based on the analysis result, including:

[0035] Analyze user actions and generate solutions based on the analysis results, including:

[0036] Perform sentiment analysis on user actions to obtain sentiment vectors of several users;

[0037] Analyze the interaction intention of user actions and determine the user's interaction intention data based on historical interaction data;

[0038] Determine the user's emotional vector, user's interaction intention and analysis results as parameters of user needs;

[0039] The user's demand coefficient is determined based on the parameters of the user's demand, and then a solution is generated in combination with a preset coefficient-solution database.

[0040] The present invention provides an active AI digital human interaction system based on visual analysis, comprising:

[0041] Model building module: batch extract image materials from the preset data source, extract features from each batch of image materials, determine training detection parameters based on the extracted features, and then determine the detection model;

[0042] Image analysis module: collects images of the target area in real time based on preset acquisition equipment, analyzes the images in combination with the detection model, and determines whether to execute the interactive process based on the image analysis results;

[0043] Data collection module: executes the interactive process and obtains user consultation information for analysis. At the same time, it collects user actions in real time based on preset collection equipment and detection models;

[0044] Action analysis module: analyzes user actions and generates solutions based on the analysis results;

[0045] Farewell execution module: obtains real-time image analysis results and determines whether to execute the farewell process.

[0046] Compared with the prior art, the present invention has the following beneficial effects:

[0047] Through visual analysis combined with detection models, active AI digital human interaction is achieved, which solves the problems of passive interaction mode, delayed response and single-dimensional analysis in existing technologies. By accurately extracting image features, real-time picture and user action analysis, and dynamically judging interaction and farewell processes, the intelligence, real-time nature and user experience of the interaction are significantly improved, ensuring the smoothness and integrity of the interaction process. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0049] Figure 1 It is a flowchart of an active AI digital human interaction method based on visual analysis provided by an embodiment of the present invention.

[0050] Figure 2 It is a structural schematic diagram of an active AI digital human interaction system based on visual analysis provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0051] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0052] Embodiment 1:

[0053] The embodiment of the present invention provides an active AI digital human interaction method based on visual analysis, such as Figure 1 As shown, including:

[0054] Step 1: Batch extract image materials from a preset data source, perform feature extraction on each batch of image materials, determine training detection parameters based on the extracted features, and then determine the detection model;

[0055] Step 2: Based on the preset acquisition equipment, the images of the target area are collected in real time, and the images are analyzed in combination with the detection model. Based on the image analysis results, it is determined whether to execute the interactive process;

[0056] Step 3: Execute the interactive process and obtain the user's consultation information for analysis. At the same time, collect user actions in real time based on the preset collection equipment and detection model;

[0057] Step 4: Analyze the user's actions and generate a solution based on the analysis results;

[0058] Step 5: Obtain the real-time image analysis results, and then determine whether to execute the farewell process.

[0059] In this embodiment, the preset data sources include static and dynamic image data, video data, scene data and interference sample data.

[0060] In this embodiment, the training detection parameters are a set of core parameters for model optimization generated by extracting and analyzing features of a preset data source.

[0061] The beneficial effects of the above technical solution: through visual analysis combined with detection models, active AI digital human interaction is realized, overcoming the problems of passive interaction, delayed response and single analysis in existing technologies. Through feature extraction and dynamic detection, user emotions and actions are accurately identified, and interaction strategies are adjusted in real time to improve the naturalness and intelligence of interaction. Combined with data analysis, full-process intelligent judgment from interaction start to farewell is realized, significantly improving the real-time, accuracy and user satisfaction of interaction, adapting to the needs of multiple scenarios, and ensuring that the interaction process is efficient, smooth and personalized.

[0062] Embodiment 2:

[0063] The embodiment of the present invention provides an active AI digital human interaction method based on visual analysis, which extracts batches of image materials from a preset data source, extracts features from each batch of image materials, determines training detection parameters based on the extracted features, and then determines a detection model, including:

[0064] Batch extracting image materials from a preset data source, classifying and labeling each batch of image materials, and then determining several groups of image materials;

[0065] Extracting a plurality of preset types of features from each group of image materials based on a first preset algorithm, and then determining a plurality of first feature sets;

[0066] Expanding each first feature set based on a second preset algorithm, thereby determining a plurality of second feature sets;

[0067] Extracting a plurality of third features from each second feature set based on a preset learning algorithm, and then determining the third features corresponding to each second feature set as a third feature set;

[0068] All third feature sets are used as training detection parameters, and a detection model is trained based on the training detection parameters, thereby determining the detection model.

[0069] In this embodiment, the first preset algorithm is used to extract several preset types of features from each group of image materials. It is the initial stage of feature extraction, focusing on capturing basic and clear features in the image, for example, using an edge detection algorithm (such as Canny edge detection) to extract contour features in the image.

[0070] In this embodiment, the second preset algorithm is used to expand the first feature set to generate a higher dimensional and richer feature set. For example, the SIFT (Scale-Invariant Feature Transform) algorithm is used to extract local key point features, and rotation and scaling transformations are applied to expand the data. Application scenario: for the same set of image materials, an extended feature set is generated under multiple angles and different lighting conditions.

[0071] In this embodiment, the preset learning algorithm is used to select key features from the second feature set and form the final third feature set. This step is a process of feature simplification and optimization to ensure that the features of the training model are efficient and representative. Features: Use machine learning methods (such as principal component analysis and feature selection algorithms) to select features, extract features with high discrimination ability, and reduce the interference of irrelevant information. For example: Algorithm type: Principal component analysis (PCA): Reduce the dimension of high-dimensional feature sets and extract the most important principal components. Recursive feature elimination (RFE): Through repeated training models, select the features with the greatest contribution. Application scenario: For the second feature set containing texture, color and shape features, use PCA to extract the key features that can best distinguish user actions (such as waving and nodding) to form the third feature set.

[0072] The beneficial effects of the above technical solution are: by batch extracting image materials, extracting and expanding feature sets layer by layer, and finally generating efficient training detection parameters, the performance of the detection model is optimized. Compared with the existing technology, this method solves the problems of insufficient image feature extraction and low model detection accuracy, and can effectively improve the integrity of feature extraction and the detection accuracy of the model, and is suitable for active AI digital human interaction in complex scenes.

[0073] Embodiment 3:

[0074] The embodiment of the present invention provides an active AI digital human interaction method based on visual analysis, which batch extracts image materials from a preset data source, including:

[0075] Evaluate the image material of each preset data source to determine the stratification coefficient of each image material of each preset data source;

[0076] Determine the data level corresponding to each image material of each preset data source based on the hierarchical coefficient of each image material of each preset data source and the preset coefficient-level data table, and then divide the image material of each preset data source into a plurality of data levels;

[0077] Determine a number of sampling windows based on each data level of each preset data source and the quantum random number generator, and determine the scale of each sampling window;

[0078] Determine the size of the acquisition unit of each sampling window of each data level of each preset data source based on the scale of the sampling window of each data level of each preset data source and a preset scale-unit database;

[0079] The image materials of each data level of each preset data source are batch extracted based on the size of the acquisition unit of each sampling window of each data level of each preset data source.

[0080] In this embodiment, each image material is divided into a specific data level (such as high level, medium level, and low level) according to a preset coefficient-level data table by evaluating its characteristics (such as clarity, resolution, color complexity, etc.). Materials at different levels may have different importance or analysis depth. For example, assuming that there is a preset data source containing multiple image materials:

[0081] Image A: high resolution, rich details (high stratification coefficient), classified as a high-level data layer; Image B: medium resolution, moderate amount of information, classified as a medium-level data layer; Image C: low resolution, blurred content, classified as a low-level data layer.

[0082] In this embodiment, the sampling window is the area in the image material used to extract data, and its scale (size) is determined by the data level. The sampling window for high-level data is generally larger, and the sampling window for low-level data is smaller to capture detailed information at different levels. For example:

[0083] For image A at a high data level, the sampling window may be a 100×100 pixel area. For image B at a medium data level, the sampling window may be a 50×50 pixel area. For image C at a low data level, the sampling window may be a 25×25 pixel area.

[0084] In this embodiment, the acquisition unit is a smaller basic data unit in the sampling window, which is used to refine the data extraction process. The size of the acquisition unit is usually further determined according to the scale of the sampling window and the image level. For example: for image A at a high-level data level, the sampling window is 100×100 pixels, and the acquisition unit may be a block of 10×10 pixels. For image B at a medium-level data level, the sampling window is 50×50 pixels, and the acquisition unit may be a block of 5×5 pixels. For image C at a low-level data level, the sampling window is 25×25 pixels, and the acquisition unit may be a block of 2×2 pixels.

[0085] The beneficial effects of the above technical solution are as follows: by dividing the image data levels based on the stratification coefficient and combining the quantum random number generator to dynamically determine the sampling window and acquisition unit size, efficient batch extraction of image materials can be achieved. Compared with the existing technology, this method solves the problems of low efficiency and insufficient stratification accuracy in the traditional image data extraction process, and significantly improves the flexibility of data processing and the accuracy of the interactive model.

[0086] Embodiment 4:

[0087] The embodiment of the present invention provides an active AI digital human interaction method based on visual analysis, which evaluates the image material of each preset data source and determines the hierarchical coefficient of each image material of each preset data source, including: ;

[0088] in, is the layering coefficient of the i-th image material of the u-th preset data source, is the total number of image features of the i-th image material of the u-th preset data source, is the g-th eigenvalue of the i-th image material of the u-th preset data source, is the feature mean of the i-th image material of the u-th preset data source, is the weight of the g-th feature of the i-th image material of the u-th preset data source, is the dynamic adjustment factor of the g-th feature of the i-th image material of the u-th preset data source, is the preset exponential magnification factor, is the preset smoothing factor, is the preset regularization factor.

[0089] In this embodiment, the dynamic adjustment factor is used to adjust the influence of the characteristic value of the image material on the layering coefficient, and its role is to dynamically enhance or weaken the importance of certain features according to specific scenes or needs, thereby optimizing the accuracy of image layering. This factor is usually set according to real-time environmental parameters, user needs or specific algorithms to ensure that the results are more flexible and adaptable.

[0090] The beneficial effects of the above technical solution are as follows: by comprehensively considering parameters such as feature weights, dynamic adjustment factors and regularization, the calculation method of the stratification coefficient is optimized to achieve accurate evaluation and batch extraction of image materials. Compared with the existing technology, it solves the problems of inflexible feature selection and low stratification accuracy, and significantly improves data processing efficiency and the intelligence level of active AI digital human interaction.

[0091] Embodiment 5:

[0092] The embodiment of the present invention provides an active AI digital human interaction method based on visual analysis, which collects images of a target area in real time based on a preset collection device, analyzes the images in combination with a detection model, and determines whether to execute an interaction process based on the image analysis result, including:

[0093] The target area is locked based on the dynamic area detection algorithm, and the image of the target area is captured in real time based on the preset acquisition equipment;

[0094] Pre-process the captured images, and then analyze the captured images frame by frame based on the detection model to obtain the collected user information and determine whether the user actively sends an interactive signal;

[0095] Processing the collected user information to determine the amount of user information;

[0096] If it is detected that the user actively sends an interactive signal or the amount of user information is greater than or equal to the preset information threshold, the interactive process will be entered. If no interactive signal is detected or the amount of user information is less than the preset information threshold, the screen will continue to be monitored.

[0097] In this embodiment, the dynamic area detection algorithm is a technology for real-time analysis of image changes to automatically lock the target area (such as a user or object). It quickly determines the area of ​​interest by analyzing the movement, color changes or distribution of feature points in the image, reduces interference from irrelevant information, and improves detection efficiency. For example, in a surveillance scene, the dynamic area detection algorithm can detect moving objects in the image in real time, such as a person walking in a room. The algorithm locks the person's position as the target area and ignores static backgrounds such as furniture and walls.

[0098] In this embodiment, the preset information volume threshold is a standard used by the system to judge the richness of information in the picture. It indicates the minimum amount of information (such as the number, complexity or quality of features) required to decide whether to continue the interactive process. For example, assuming the threshold is 10 feature points, if only a static outline is detected in the picture, the information volume is 5, which is lower than the threshold, then the system continues to monitor. If a clear face with rich facial expressions is detected, the information volume reaches 15, then the threshold requirement is met and the interactive process is entered.

[0099] In this embodiment, the interactive process is a behavioral pattern that is initiated when the system detects that a condition is met (such as an active signal or sufficient information), usually including actions, conversations or responses to interact with the user. For example: in the AI ​​digital human system in a shopping mall, when the user waves (active signal) or the system detects a clear face and voice (meets the information threshold), the system starts the interactive process, such as greeting, recommending products or answering user questions.

[0100] The beneficial effects of the above technical solution are as follows: through the dynamic area detection algorithm and frame-by-frame analysis technology, the target area can be accurately locked and user information can be obtained in real time. The interactive signal and information volume threshold are combined to determine whether to enter the interactive process. This solves the problems of low interaction efficiency and high false trigger rate in the existing technology, and realizes efficient and intelligent active AI digital human interaction.

[0101] Embodiment 6:

[0102] The embodiment of the present invention provides an active AI digital human interaction method based on visual analysis, which processes the collected user information and further determines the amount of user information, including:

[0103] Extract parameters from the collected user information, and then obtain relevant parameters of the user information volume;

[0104] The amount of user information is determined based on parameters related to the amount of user information, including: ;in, is the amount of user information, N is the number of users in the screen, The time frame number of action collection, For the The key points of the user's body in the frame, For the The key points of the user's body in the frame, is the number of user expression features in the picture, is the intensity value of the jth user’s facial expression feature, is the preset user motion amplitude conversion coefficient, It is the preset user expression intensity conversion coefficient.

[0105] In this embodiment, the key point position of the user's body refers to the coordinate information used to describe the main joints or parts of the human body in human posture recognition. These key points usually include the head, shoulders, elbows, wrists, hips, knees, ankles, etc., which can reflect the user's movements or postures. By analyzing the position changes of these key points, the user's movement amplitude and behavioral intentions can be judged. For example: In the dance training system, AI identifies the key point positions of the dancer through visual capture. For example, the key point coordinates of the left shoulder in the 5th frame are (200, 300) and the right shoulder is (250, 300). Over time, the coordinates of the left shoulder in the 6th frame become (220, 280), and the right shoulder becomes (270, 280). The system analyzes the movement trajectory of the dancer leaning to the right.

[0106] The beneficial effects of the above technical solution are as follows: by extracting the positions of key points of the user's body and facial expression features, combining the movement amplitude and expression intensity conversion coefficient, the amount of user information can be accurately quantified, which solves the problems of insufficient comprehensive analysis of user behavior and emotion and insufficient interaction accuracy in the existing technology, and improves the intelligence and real-time nature of AI digital human interaction.

[0107] Embodiment 7:

[0108] The embodiment of the present invention provides an active AI digital human interaction method based on visual analysis, which analyzes user actions and then generates a solution based on the analysis results, including:

[0109] Analyze user actions and generate solutions based on the analysis results, including:

[0110] Perform sentiment analysis on user actions to obtain sentiment vectors of several users;

[0111] Analyze the interaction intention of user actions and determine the user's interaction intention data based on historical interaction data;

[0112] Determine the user's emotional vector, user's interaction intention and analysis results as parameters of user needs;

[0113] The user's demand coefficient is determined based on the parameters of the user's demand, and then a solution is generated in combination with a preset coefficient-solution database.

[0114] In this embodiment, the first analysis refers to a preliminary analysis of the user's actions to extract emotion-related features in the actions, such as body language, facial expressions, etc., for generating an emotion vector. For example, when a user smiles or nods slightly through the camera, the system recognizes the emotion states such as "friendly" and "happy" through the first analysis.

[0115] In this embodiment, the emotion vector is multidimensional data used to represent the user's current emotional state. Each dimension corresponds to a different emotional feature (such as happiness, anger, sadness, etc.), which can quantify the user's emotions. For example, if the user shows a smile and a relaxed posture, the emotion vector may be represented as [0.8 (happy), 0.2 (calm), 0.0 (angry)].

[0116] In this embodiment, the second analysis refers to identifying the user's specific interaction intention by further analyzing the user's actions and combining historical interaction data (such as the user's previous behaviors and preferences). For example, if a user points a finger at an object and tilts his head to look at the camera, combined with the user's previous question and answer records, the second analysis determines that the user may be asking "what is this".

[0117] In this embodiment, the user's demand coefficient is determined based on the parameters of the user's demand. The user's demand coefficient is calculated comprehensively based on the emotion vector, interaction intention and other analysis parameters, and is used to measure the urgency and clarity of the user's current demand. For example, the system analyzes the user's quick gestures, serious expression (the emotion vector shows anxiety 0.9), and voice inquiry "Tell me what the weather is like", and generates a high demand coefficient of 0.95, prompting the system to give priority to answering.

[0118] The beneficial effects of the above technical solution are as follows: through multi-layer analysis of user actions, emotion vectors and interaction intentions are extracted, historical data are combined to generate user demand parameters and demand coefficients, and solutions are accurately matched. This solves the problems of inaccurate recognition of user emotions and intentions and unintelligent responses in the existing technology, and improves the naturalness and effectiveness of AI digital human interaction.

[0119] Embodiment 8:

[0120] The embodiment of the present invention provides an active AI digital human interaction method based on visual analysis, such as Figure 2 As shown, including:

[0121] Model building module: batch extract image materials from the preset data source, extract features from each batch of image materials, determine training detection parameters based on the extracted features, and then determine the detection model;

[0122] Image analysis module: collects images of the target area in real time based on preset acquisition equipment, analyzes the images in combination with the detection model, and determines whether to execute the interactive process based on the image analysis results;

[0123] Data collection module: executes the interactive process and obtains user consultation information for analysis. At the same time, it collects user actions in real time based on preset collection equipment and detection models;

[0124] Action analysis module: analyzes user actions and generates solutions based on the analysis results;

[0125] Farewell execution module: obtains real-time image analysis results and determines whether to execute the farewell process.

[0126] The beneficial effects of the above technical solution are as follows: through visual analysis combined with the detection model, active AI digital human interaction is realized, which solves the problems of passive interaction mode, delayed response and single-dimensional analysis in the existing technology. By accurately extracting image features, real-time picture and user action analysis, and dynamically judging the interaction and farewell process, the intelligence, real-time nature and user experience of the interaction are significantly improved, ensuring the smoothness and integrity of the interaction process.

[0127] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An active AI digital human interaction method based on visual analysis, characterized in that: include: Step 1: Batch extract image materials from a preset data source, perform feature extraction on each batch of image materials, determine training detection parameters based on the extracted features, and then determine the detection model; Step 2: Based on the preset acquisition equipment, the images of the target area are collected in real time, and the images are analyzed in combination with the detection model. Based on the image analysis results, it is determined whether to execute the interactive process; Step 3: Execute the interactive process and obtain the user's consultation information for analysis. At the same time, collect user actions in real time based on the preset collection equipment and detection model; Step 4: Analyze the user's actions and generate a solution based on the analysis results; Step 5: Obtain the real-time image analysis results, and then determine whether to execute the farewell process.

2. The active AI digital human interaction method based on visual analysis according to claim 1, characterized in that: Batch extract the image materials of the preset data source, extract features from each batch of image materials, determine the training detection parameters based on the extracted features, and then determine the detection model, including: Batch extracting image materials from a preset data source, classifying and labeling each batch of image materials, and then determining several groups of image materials; Extracting a plurality of preset types of features from each group of image materials based on a first preset algorithm, and then determining a plurality of first feature sets; Expanding each first feature set based on a second preset algorithm, thereby determining a plurality of second feature sets; Extracting a plurality of third features from each second feature set based on a preset learning algorithm, and then determining the third features corresponding to each second feature set as a third feature set; All third feature sets are used as training detection parameters, and a detection model is trained based on the training detection parameters, thereby determining the detection model.

3. The active AI digital human interaction method based on visual analysis according to claim 2, characterized in that: Batch extract image materials from preset data sources, including: Evaluate the image material of each preset data source to determine the stratification coefficient of each image material of each preset data source; Determine the data level corresponding to each image material of each preset data source based on the hierarchical coefficient of each image material of each preset data source and the preset coefficient-level data table, and then divide the image material of each preset data source into a plurality of data levels; Determine a number of sampling windows based on each data level of each preset data source and the quantum random number generator, and determine the scale of each sampling window; Determine the size of the acquisition unit of each sampling window of each data level of each preset data source based on the scale of the sampling window of each data level of each preset data source and a preset scale-unit database; The image materials of each data level of each preset data source are batch extracted based on the size of the acquisition unit of each sampling window of each data level of each preset data source.

4. The active AI digital human interaction method based on visual analysis according to claim 3, characterized in that: The image material of each preset data source is evaluated to determine the stratification coefficient of each image material of each preset data source, including: ； in, is the layering coefficient of the i-th image material of the u-th preset data source, is the total number of image features of the i-th image material of the u-th preset data source, is the g-th eigenvalue of the i-th image material of the u-th preset data source, is the feature mean of the i-th image material of the u-th preset data source, is the weight of the g-th feature of the i-th image material of the u-th preset data source, is the dynamic adjustment factor of the g-th feature of the i-th image material of the u-th preset data source, is the preset exponential magnification factor, is the preset smoothing factor, is the preset regularization factor.

5. The active AI digital human interaction method based on visual analysis according to claim 1, characterized in that: Based on the preset acquisition equipment, the images of the target area are collected in real time, and the images are analyzed in combination with the detection model. Based on the image analysis results, it is determined whether to execute the interactive process, including: The target area is locked based on the dynamic area detection algorithm, and the image of the target area is captured in real time based on the preset acquisition equipment; Pre-process the captured images, and then analyze the captured images frame by frame based on the detection model to obtain the collected user information and determine whether the user actively sends an interactive signal; Processing the collected user information to determine the amount of user information; If it is detected that the user actively sends an interactive signal or the amount of user information is greater than or equal to the preset information threshold, the interactive process will be entered. If no interactive signal is detected or the amount of user information is less than the preset information threshold, the screen will continue to be monitored.

6. The active AI digital human interaction method based on visual analysis according to claim 5, characterized in that: Process the collected user information to determine the amount of user information, including: Extract parameters from the collected user information, and then obtain relevant parameters of the user information volume; The amount of user information is determined based on parameters related to the amount of user information, including: ； in, is the amount of user information, N is the number of users in the screen, is the number of time frames for action collection, is the key point position of the user's body in the kth frame, is the key point position of the user's body in the k+1th frame, is the number of user expression features in the picture, is the intensity value of the jth user's expression feature, A1 is the preset user action amplitude conversion coefficient, It is the preset user expression intensity conversion coefficient.

7. The active AI digital human interaction method based on visual analysis according to claim 1, characterized in that: Analyze user actions and generate solutions based on the analysis results, including: Perform sentiment analysis on user actions to obtain sentiment vectors of several users; Analyze the interaction intention of user actions and determine the user's interaction intention data based on historical interaction data; Determine the user's emotional vector, user's interaction intention and analysis results as parameters of user needs; The user's demand coefficient is determined based on the parameters of the user's demand, and then a solution is generated in combination with a preset coefficient-solution database.

8. An active AI digital human interaction system based on visual analysis, applied to an active AI digital human interaction method based on visual analysis as claimed in any one of claims 1 to 7, characterized in that: include: Model building module: batch extract image materials from the preset data source, extract features from each batch of image materials, determine training detection parameters based on the extracted features, and then determine the detection model; Image analysis module: collects images of the target area in real time based on preset acquisition equipment, analyzes the images in combination with the detection model, and determines whether to execute the interactive process based on the image analysis results; Data collection module: executes the interactive process and obtains user consultation information for analysis. At the same time, it collects user actions in real time based on preset collection equipment and detection models; Action analysis module: analyzes user actions and generates solutions based on the analysis results; Farewell execution module: obtains real-time image analysis results and determines whether to execute the farewell process.

Images