Intelligent AI-driven home scene adaptive adjustment method and system

Abstract

The invention relates to the field of artificial intelligence, and discloses an intelligent AI-driven home scene adaptive adjustment method and system, and the method comprises the steps: collecting the environment data of a home environment, and the video image data and face temperature data of a home user in the home environment; the method comprises the steps of determining environment data of a home user, determining behavior data of the home user, analyzing facial expression data of the home user, performing data fusion on the environment data, the behavior data and the facial expression data, determining a high-frequency action path and an emotional intention of the home user, analyzing an equipment parameter distribution state of home equipment in a home environment, and determining personalized home preference of the home user; constructing a home equipment parameter control scheme of the home user, and performing adaptive adjustment on the home environment to obtain an adjusted home environment; and analyzing the comfort level of the home user for the current home environment, optimizing and adjusting the home environment, and obtaining a target home environment. According to the invention, the accuracy of home scene adaptive adjustment can be improved.

Description

Technical Field

[0001] The present invention relates to an adaptive adjustment method and system for home scenes driven by intelligent AI, and belongs to the field of artificial intelligence. Background Art

[0002] The home scene driven by intelligent AI refers to the use of artificial intelligence (AI) technology, especially advanced algorithms such as machine learning, deep learning, and natural language processing, combined with various sensors and actuators to achieve intelligent perception, decision-making, and control of the home environment, thereby providing users with a more comfortable, convenient, safe, and energy-saving living experience.

[0003] While traditional smart home systems can implement some basic automation functions, such as timed lighting and temperature-controlled air conditioning, they lack real-time perception and deep understanding of user behavior and environmental changes. This makes it difficult for them to flexibly and accurately adapt to the individual needs of different users and dynamically changing environmental factors. Existing smart home systems struggle to fully and accurately capture this complex and ever-changing information and respond promptly and effectively, resulting in a less-than-ideal user experience. Summary of the Invention

[0004] The present invention provides a method and system for adaptively adjusting home scenes under intelligent AI drive, the main purpose of which is to improve the accuracy of adaptive adjustment of home scenes.

[0005] To achieve the above objectives, the present invention provides an intelligent AI-driven adaptive adjustment method for home scenes, comprising:

[0006] Configuring multiple sensors within a home environment, and collecting environmental data of the home environment and video image data and facial temperature data of home users within the home environment based on the multiple sensors;

[0007] extracting behavioral features from the video image data, determining behavioral data of the household user based on the behavioral features, analyzing facial expression data of the household user based on the facial temperature data and the video image data, and fusing the environmental data, the behavioral data, and the facial expression data using a data fusion module of a preset intelligent AI analysis model to obtain fused data;

[0008] Based on the fused data, the high-frequency action paths and emotional intentions of the home user are determined, and the device parameter distribution status of the home devices in the home environment is analyzed. Based on the high-frequency action paths, the spatial preference matrix of the home user is determined. Based on the emotional intention, the psychological state map of the home user is analyzed. Based on the device parameter distribution status, the functional distribution requirements of the home user are analyzed. Based on the spatial preference matrix, the psychological state map, and the functional distribution requirements, the preference fusion engine of the preset intelligent AI analysis model is used to determine the personalized home preferences of the home user.

[0009] Based on the personalized home preferences, a home device parameter control scheme is constructed for the home user, real-time home information of the home user and real-time environmental information of the home environment are collected, and the home environment is adaptively adjusted using the home device parameter control scheme in combination with the real-time home information and the real-time environmental information to obtain an adjusted home environment;

[0010] The expression changes of the home user in the adjusted home environment are collected in real time, and based on the expression changes, the comfort level of the home user with the current home environment is analyzed. Based on the comfort level, the adjusted home environment is optimized to obtain a target home environment.

[0011] Optionally, extracting behavioral features of the video image data includes:

[0012] Performing filtering and denoising processing on the video image data to obtain denoised video image data;

[0013] Identifying sensitive areas in the denoised video image data, and performing pixelation desensitization processing on the denoised video image data according to the sensitive areas to obtain desensitized video image data;

[0014] Identifying a target object of the desensitized video image data and generating a virtual target object of the target object;

[0015] Determining a virtual character position of the virtual target object in consecutive frames;

[0016] Based on the position of the virtual character, detecting key points of the body of the virtual target object;

[0017] Calculating body motion parameters of the virtual target object based on the body key points;

[0018] Extracting posture features and gesture features of the virtual target object;

[0019] The behavior characteristics of the virtual target object are determined according to the body movement parameters, the posture characteristics, and the gesture characteristics.

[0020] Optionally, analyzing the facial expression data of the household user based on the facial temperature data and the video image data includes:

[0021] Identifying a target facial image of the video image data and generating a virtual facial image of the target facial image;

[0022] Extracting facial feature points of the virtual face image;

[0023] Extracting expression features of the virtual face image based on the facial feature points;

[0024] determining a facial temperature distribution of the household user according to the facial temperature data;

[0025] Identifying abnormal temperature areas on the face of the household user;

[0026] determining a facial temperature feature of the household user according to the facial temperature distribution and the facial temperature abnormality area;

[0027] Classifying the expressions of the household user according to the expression features and the facial temperature features to obtain basic emotions;

[0028] Detecting slight changes in facial expressions of the household user, and determining facial expression data of the household user based on the slight changes in facial expressions and the basic emotions.

[0029] Optionally, the data fusion module using the preset intelligent AI analysis model performs data fusion on the environmental data, the behavioral data, and the facial expression data to obtain fused data, including:

[0030] Analyzing the data association relationship among the environmental data, the behavioral data, and the facial expression data;

[0031] extracting sensitive data of the environmental data, the behavioral data, and the facial expression data;

[0032] Determining a sensitivity level of the sensitive data, and constructing a hierarchical dynamic encryption algorithm for the sensitive data based on the sensitivity level;

[0033] Constructing a feature vector matrix of the environmental data, the behavioral data, and the facial expression data according to the data association relationship and the hierarchical dynamic encryption algorithm;

[0034] Determining a feature weight matrix of the feature vector matrix;

[0035] According to the eigenvector matrix and the eigenweight matrix, a data fusion algorithm in the data fusion module is used to calculate a fusion eigenvector of the environmental data, the behavioral data, and the facial expression data, wherein the data fusion algorithm includes:

[0036]

[0037] Among them, R represents the fused feature vector, e represents the exponential function with e as the base, and T h represents the environmental eigenvector in the eigenvector matrix, q h represents the environmental feature weight in the feature weight matrix, T x represents the behavioral eigenvector in the eigenvector matrix, q x represents the behavioral feature weight in the feature weight matrix, T b represents the expression eigenvector in the eigenvector matrix, T b represents the weight of the expression feature in the feature weight matrix, and C represents a constant;

[0038] According to the fused feature vector, the data fusion module is utilized to output fused data of the environmental data, the behavioral data, and the facial expression data.

[0039] Optionally, determining the household user's high-frequency action path and emotional intention based on the fused data includes:

[0040] constructing an action grid of the household user based on the fused data;

[0041] Constructing a transition probability matrix of the household user according to the action grid;

[0042] According to the transition probability matrix, the high-frequency action path of the household user is calculated using the following formula:

[0043]

[0044] Among them, PathScore(G) represents the high-frequency action path, G represents the movement path of the home user, n represents the number of grids in the action grid, and D gigi+1 Represents the transition probability matrix from the action grid g i Move to action grid g i+1 The probability of moving path G is , e represents the exponential function with base e, u represents the hyperparameter, and t(G) represents the total time taken to move path G;

[0045] Extracting emotional features of the fused data, and constructing an emotional intention analysis model of the home user based on the emotional features;

[0046] Based on the fusion data, the emotional intention of the home user is analyzed using the emotional intention analysis model.

[0047] Optionally, determining the spatial preference matrix of the household user based on the high-frequency action path includes:

[0048] Encoding the high-frequency motion path to obtain an encoded trajectory;

[0049] Virtually constructing a spatial grid of the household user;

[0050] Mapping the encoded trajectory into a spatial grid to calculate the number of user visits and user stay time of the spatial grid;

[0051] Determine the spatial preference matrix of the home user according to the number of user visits and the user's stay time.

[0052] Optionally, the determining of the personalized home preferences of the home user based on the spatial preference matrix, the psychological state map, and the functional distribution requirements using a preference fusion engine of a preset intelligent AI analysis model includes:

[0053] defining a preference scoring function of the preference fusion engine;

[0054] Analyzing the spatial preference matrix, the mental state map, and the spatial preference weights, mental state weights, and functional requirement weights of the functional distribution requirements;

[0055] Determining the number of spatial regions, psychological dimensions, and functions of the spatial preference matrix, the psychological state map, and the functional distribution requirements;

[0056] Calculating the personalized home preference score of the home user using the preference score function according to the spatial preference weight, the psychological state weight, the functional requirement weight, the number of spatial areas, the psychological dimension, and the number of functions;

[0057] Based on the personalized home preference score, the personalized home preference of the home user is determined.

[0058] Optionally, the utilizing the personalized control solution for home devices in combination with the real-time home information and the real-time environment information to adaptively adjust the home environment to obtain an adjusted home environment includes:

[0059] Determining a current user of the home environment based on the real-time home information;

[0060] Understanding the personalized service needs of the current user based on the personalized home preferences corresponding to the current user and the real-time environmental information;

[0061] Analyze changes in demand for personalized service needs in real time, and determine dynamic adjustment parameters of home devices corresponding to the home environment based on the changes in demand;

[0062] Determining the control logic of the home device according to the personalized control solution of the home device;

[0063] According to the control logic and the dynamic adjustment parameters, the home environment is adaptively adjusted to obtain an adjusted home environment.

[0064] Optionally, analyzing the comfort level of the home user with the current home environment based on the expression change includes:

[0065] Determining a current activity of the home user, and analyzing the expression meaning of the expression change based on the current activity;

[0066] Calculating the expression intensity and expression duration of the expression change based on the expression meaning;

[0067] Classifying the expression changes according to the expression intensity and the expression duration to obtain multiple categories of expressions, wherein the multiple categories of expressions include: positive expressions, neutral expressions, and negative expressions;

[0068] Constructing a comfort mapping table of the home user to the current home environment according to the multiple types of expressions;

[0069] The expression change is mapped to the comfort mapping table to obtain the comfort level of the home user with the current home environment.

[0070] In order to solve the above problems, the present invention also provides a home scene adaptive adjustment system driven by intelligent AI, the system comprising:

[0071] a data acquisition module, configured to configure multiple sensors within a home environment, and based on the multiple sensors, collect environmental data of the home environment and video image data and facial temperature data of home users within the home environment;

[0072] a data fusion module, configured to extract behavioral features from the video image data, determine behavioral data of the household user based on the behavioral features, analyze facial expression data of the household user based on the facial temperature data and the video image data, and fuse the environmental data, the behavioral data, and the facial expression data using a data fusion module of a preset intelligent AI analysis model to obtain fused data;

[0073] a personalized home preference analysis module for determining the home user's high-frequency action paths and emotional intentions based on the fused data, analyzing the device parameter distribution status of home devices in the home environment, determining the home user's spatial preference matrix based on the high-frequency action paths, analyzing the home user's psychological state map based on the emotional intentions, analyzing the home user's functional distribution requirements based on the device parameter distribution status, and determining the home user's personalized home preferences based on the spatial preference matrix, the psychological state map, and the functional distribution requirements using a preference fusion engine of a preset intelligent AI analysis model;

[0074] a home environment adjustment module, configured to construct a home device parameter control scheme for the home user based on the personalized home preferences, collect real-time home information of the home user and real-time environmental information of the home environment, and adaptively adjust the home environment using the home device parameter control scheme in combination with the real-time home information and the real-time environmental information to obtain an adjusted home environment;

[0075] The target home environment determination module is used to collect the changes in the home user's expression in the adjusted home environment in real time, analyze the home user's comfort with the current home environment based on the changes in expression, and optimize the adjusted home environment based on the comfort to obtain the target home environment.

[0076] Compared with the problems described in the background technology, the embodiment of the present invention can realize comprehensive perception of the home environment by collecting the environmental data of the home environment and the video image data and facial temperature data of the home users in the home environment based on the multi-sensor, thereby providing sufficient data sources for subsequent adaptive adjustment of the home scene; optionally, the embodiment of the present invention uses the data fusion module of the preset intelligent AI analysis model to fuse the environmental data, the behavioral data and the facial expression data, and the obtained fused data can more accurately analyze and meet the needs of the user, thereby improving the user experience; the embodiment of the present invention determines the personalized home preferences of the home users based on the spatial preference matrix, the psychological state map and the functional distribution requirements, and the preference fusion engine can integrate this information to generate a personalized home Home settings and automation scenes, thereby providing a customized comfort experience; the embodiment of the present invention utilizes the personalized control scheme of the home equipment in combination with the real-time home information and the real-time environmental information to adaptively adjust the home environment, and obtains that the adjusted home environment can significantly improve the comfort, safety, health and energy efficiency of the living environment, while achieving a higher level of automation and personalized services; the embodiment of the present invention can more accurately judge whether the current environmental adjustment has met the user's expectations and whether further adjustment is needed by collecting the changes in the expression of the home user in the adjusted home environment in real time; finally, the embodiment of the present invention optimizes the adjusted home environment based on the comfort level to obtain the target home environment. By accurately understanding the user's comfort level, it is possible to create a highly personalized environment to meet the unique needs of each user, thereby significantly improving the comfort and satisfaction of living. Therefore, the method and system for adaptive adjustment of home scenes driven by intelligent AI provided by the embodiment of the present invention can improve the accuracy of adaptive adjustment of home scenes. BRIEF DESCRIPTION OF THE DRAWINGS

[0077] Figure 1 A schematic diagram of a flow chart of a method for adaptively adjusting home scenes driven by intelligent AI according to an embodiment of the present invention;

[0078] Figure 2 A schematic diagram of a module for implementing the adaptive adjustment method for home scenes driven by intelligent AI, provided in one embodiment of the present invention.

[0079] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0080] It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0081] The embodiment of the present application provides a method for adaptively adjusting a home scene under intelligent AI drive. The execution subject of the method for adaptively adjusting a home scene under intelligent AI drive includes but is not limited to at least one of the electronic devices such as a server and a terminal that can be configured to execute the method provided by the embodiment of the present application. In other words, the method for adaptively adjusting a home scene under intelligent AI drive can be executed by software or hardware installed on a terminal device or a server device. The server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, etc.

[0082] Example 1:

[0083] Reference Figure 1 FIG2 is a flow chart of a method for adaptively adjusting a home scene under intelligent AI driving according to an embodiment of the present invention. In this embodiment, the method for adaptively adjusting a home scene under intelligent AI driving includes:

[0084] S1. Configure multiple sensors in a home environment, and based on the multiple sensors, collect environmental data of the home environment and video image data and facial temperature data of home users in the home environment.

[0085] By configuring multiple sensors within a home environment, embodiments of the present invention can collect relevant data within the home environment, providing a data foundation for subsequent analysis. Multi-sensors refer to a combination of multiple different types of sensors, such as temperature sensors, light sensors, microwave sensors, and cameras, that work together to collect comprehensive information about the home environment and user status.

[0086] The embodiment of the present invention can achieve comprehensive perception of the home environment by collecting environmental data of the home environment and video image data and facial temperature data of home users in the home environment based on the multi-sensor, providing a sufficient data source for subsequent adaptive adjustment of home scenes. The environmental data refers to quantitative information about the state of the home environment collected by various environmental sensors. The video image data refers to visual information about people, objects and scenes in the home environment collected by image sensors (such as cameras). The facial temperature data refers to quantitative information about the facial temperature distribution of home users collected by thermal imaging sensors (such as infrared thermal imaging cameras).

[0087] S2. Extract behavioral features from the video image data, determine behavioral data of the household user based on the behavioral features, analyze facial expression data of the household user based on the facial temperature data and the video image data, and fuse the environmental data, the behavioral data, and the facial expression data using a data fusion module of a preset intelligent AI analysis model to obtain fused data.

[0088] Embodiments of the present invention can identify user behaviors by extracting behavioral features from the video image data, thereby better understanding user needs and habits and providing more personalized services. The behavioral features refer to visual information about the movements, postures, and gestures of home users collected by image sensors (such as cameras).

[0089] As an embodiment of the present invention, extracting the behavioral features of the video image data includes:

[0090] Performing filtering and denoising processing on the video image data to obtain denoised video image data;

[0091] Identifying sensitive areas in the denoised video image data, and performing pixelation desensitization processing on the denoised video image data according to the sensitive areas to obtain desensitized video image data;

[0092] Identifying a target object of the desensitized video image data and generating a virtual target object of the target object;

[0093] Determining a virtual character position of the virtual target object in consecutive frames;

[0094] Based on the position of the virtual character, detecting key points of the body of the virtual target object;

[0095] Calculating body motion parameters of the virtual target object based on the body key points;

[0096] Extracting posture features and gesture features of the virtual target object;

[0097] The behavior characteristics of the virtual target object are determined according to the body movement parameters, the posture characteristics, and the gesture characteristics.

[0098] The denoised video image data refers to video image data that has undergone filtering and denoising. Sensitive areas refer to those portions of the video that contain personal privacy information and need to be protected or hidden, such as name tags, work badges, student ID cards, special clothing, height, body shape, etc. Desensitized video image data refers to video image data that has been processed based on the original video image data by processing sensitive areas, rendering the information in these areas inaccessible to direct identification and utilization, thereby achieving the purpose of protecting privacy. The target object refers to a user in a home environment. The virtual target object refers to an abstract or digital representation created and tracked based on a real person in the original video image after pixelation and desensitization. The virtual character position refers to the spatial coordinates of the virtual target object in the image captured by an image sensor (such as a camera). Body key points refer to specific landmark locations on the human body, such as the head, torso, limbs, etc. Body motion parameters refer to quantitative indicators that describe the user's body motion state, such as motion speed, angle, displacement, etc. Posture features refer to quantitative or symbolic information that describes the user's body posture, such as joint angles, limb positions, body orientation, etc. The gesture feature refers to quantitative or symbolic information describing the user's gesture action, such as the degree of finger bending, finger extension state, palm posture, etc.

[0099] Optionally, the target object of the denoised video image data can be identified by a target detection algorithm, such as YOLO, SSD, Faster R-CNN, etc.

[0100] Optionally, the position of the target object in the continuous frames can be determined by a target tracking algorithm, such as a Kalman filter, a correlation filter, etc.

[0101] By determining the behavioral data of the home user based on the behavioral characteristics, embodiments of the present invention can identify and understand the user's ongoing activities, better understand the user's needs and intentions, and thus provide more personalized services. The behavioral data refers to various information that can reflect the user's activities and habits in the home environment.

[0102] Optionally, the behavioral data of the household user can be obtained by identifying the behavioral characteristics through a machine learning algorithm, such as a support vector machine, random forest, neural network, etc.

[0103] By analyzing the facial expression data of the household user based on the facial temperature data and the video image data, embodiments of the present invention can more accurately identify the user's emotional state and reflect the user's comfort level with the surrounding environment. The facial expression data refers to dynamic facial feature information collected and processed by a visual sensor.

[0104] As an embodiment of the present invention, analyzing the facial expression data of the household user based on the facial temperature data and the video image data includes:

[0105] Identifying a target facial image of the video image data and generating a virtual facial image of the target facial image;

[0106] Extracting facial feature points of the virtual face image;

[0107] Extracting expression features of the virtual face image based on the facial feature points;

[0108] determining a facial temperature distribution of the household user according to the facial temperature data;

[0109] Identifying abnormal temperature areas on the face of the household user;

[0110] determining a facial temperature feature of the household user according to the facial temperature distribution and the facial temperature abnormality area;

[0111] Classifying the expressions of the household user according to the expression features and the facial temperature features to obtain basic emotions;

[0112] Detecting slight changes in facial expressions of the household user, and determining facial expression data of the household user based on the slight changes in facial expressions and the basic emotions.

[0113] The target facial image refers to video image data containing facial information of a specific user. The virtual facial image refers to a new image representation generated by pixelating the original recognized target facial image. Facial feature points refer to key points with distinctive features in a facial image, such as the eyes, nose, and mouth. Expression features refer to information extracted from the user's facial feature points, used to describe and distinguish different facial expressions, such as feature point displacement, facial shape changes, and facial texture changes. The facial temperature distribution refers to information about temperature differences between different regions of the user's face. Abnormal facial temperature regions refer to areas in the facial temperature distribution that exhibit significant differences compared to normal or baseline temperatures. Facial temperature features refer to key information extracted from the facial temperature distribution and abnormal facial temperature regions, used to characterize the user's facial temperature state, such as facial temperature difference features and facial temperature gradient features. Basic emotions refer to a set of core emotions that are believed to be universal and shared across cultures and species, such as happiness, sadness, and anger. Minor expression changes refer to subtle, transient changes in the user's facial muscles.

[0114] Optionally, the target facial image of the video image data can be identified by a face detection algorithm, such as a Haar feature classifier, SSD, YOLO or a deep learning-based model.

[0115] Optionally, the facial temperature distribution can be obtained by mapping the facial temperature data into the target face image.

[0116] The embodiments of the present invention utilize a data fusion module of a preset intelligent AI analysis model to fuse the environmental data, the behavioral data, and the facial expression data. The resulting fused data can more accurately analyze and meet user needs, thereby improving the user experience. The fused data refers to data obtained by integrating multiple different types of data sources through data fusion technology.

[0117] As an embodiment of this invention, the data fusion module of the preset intelligent AI analysis model is used to fuse the environmental data, the behavioral data, and the facial expression data to obtain fused data, including:

[0118] Analyzing the data association relationship among the environmental data, the behavioral data, and the facial expression data;

[0119] extracting sensitive data of the environmental data, the behavioral data, and the facial expression data;

[0120] Determining a sensitivity level of the sensitive data, and constructing a hierarchical dynamic encryption algorithm for the sensitive data based on the sensitivity level;

[0121] Constructing a feature vector matrix of the environmental data, the behavioral data, and the facial expression data according to the data association relationship and the hierarchical dynamic encryption algorithm;

[0122] Determining a feature weight matrix of the feature vector matrix;

[0123] According to the eigenvector matrix and the eigenweight matrix, a data fusion algorithm in the data fusion module is used to calculate a fusion eigenvector of the environmental data, the behavioral data, and the facial expression data, wherein the data fusion algorithm includes:

[0124]

[0125] Among them, R represents the fused feature vector, e represents the exponential function with e as the base, and T h represents the environmental eigenvector in the eigenvector matrix, q h represents the environmental feature weight in the feature weight matrix, T x represents the behavioral eigenvector in the eigenvector matrix, qx represents the behavioral feature weight in the feature weight matrix, T b represents the expression eigenvector in the eigenvector matrix, T b represents the weight of the expression feature in the feature weight matrix, and C represents a constant;

[0126] According to the fused feature vector, the data fusion module is utilized to output fused data of the environmental data, the behavioral data, and the facial expression data.

[0127] The data association relationship refers to the inherent connections and mutual influences between environmental data, behavioral data, and facial expression data, such as correlation and causal relationships. Sensitive data refers to information that can directly or indirectly identify and reflect a user's personal characteristics, behavioral habits, physiological state, or private state, such as facial images and behavioral habits. The sensitivity level refers to a differential indicator used to measure the privacy leakage risk, potential harm level, and protection difficulty of this data, such as high sensitivity, medium sensitivity, and low sensitivity. The layered dynamic encryption algorithm refers to a comprehensive encryption algorithm that divides data into different protection levels based on attributes such as data sensitivity level, and the encryption keys, algorithms, and parameters used in these levels are dynamically adjusted and changed based on the external environment (such as time, security status, access requests) and internal policies. The eigenvector matrix refers to a matrix composed of multiple eigenvectors formed after feature extraction of environmental data, behavioral data, and facial expression data. The feature weight matrix refers to a weight coefficient matrix used to represent the importance of different features. The data fusion algorithm refers to an algorithm that integrates and processes data from multiple modalities. The fused eigenvector refers to the result obtained by multiplying and combining the eigenvectors of different modalities with their corresponding weights.

[0128] Optionally, the data association relationship among the environmental data, the behavioral data, and the facial expression data may be analyzed by a multi-layer analysis algorithm, such as a correlation analysis algorithm, a regression analysis algorithm, a cluster analysis, and the like.

[0129] Exemplarily, the sensitivity levels are high sensitivity, medium sensitivity, and low sensitivity. When the sensitive data is at the low sensitivity level, it is encrypted using the AES-128 encryption algorithm. When the sensitive data is at the medium sensitivity level, it is collaboratively encrypted using AES-256 encryption and timestamp obfuscation encryption. When the sensitive data is at the high sensitivity level, it is encrypted using the homomorphic encryption algorithm.

[0130] S3. Based on the fused data, determine the high-frequency action paths and emotional intentions of the home user, and analyze the device parameter distribution status of the home devices in the home environment; based on the high-frequency action paths, determine the spatial preference matrix of the home user; based on the emotional intentions, analyze the psychological state map of the home user; based on the device parameter distribution status, analyze the functional distribution requirements of the home user; based on the spatial preference matrix, the psychological state map and the functional distribution requirements, use the preference fusion engine of the preset intelligent AI analysis model to determine the personalized home preferences of the home user.

[0131] By determining the home user's frequent movement paths and emotional intent based on the fused data, embodiments of the present invention can analyze the user's behavioral patterns and emotional changes, automatically adjust the home environment when the user needs it, and provide more personalized services. Frequent movement paths refer to the user's frequently repeated activity patterns and movement routes in their living environment. Emotional intent refers to the user's inherent emotional state and desired emotional tendencies, inferred from analyzing the fused data.

[0132] As an embodiment of the present invention, determining the high-frequency action path and emotional intention of the household user based on the fused data includes:

[0133] constructing an action grid of the household user based on the fused data;

[0134] Constructing a transition probability matrix of the household user according to the action grid;

[0135] According to the transition probability matrix, the high-frequency action path of the household user is calculated using the following formula:

[0136]

[0137] Among them, PathScore(G) represents the high-frequency action path, G represents the movement path of the home user, n represents the number of grids in the action grid, and D gigi+1 Represents the transition probability matrix from the action grid g i Move to action grid g i+1 The probability of moving path G is , e represents the exponential function with base e, u represents the hyperparameter, and t(G) represents the total time taken to move path G;

[0138] Extracting emotional features of the fused data, and constructing an emotional intention analysis model of the home user based on the emotional features;

[0139] Based on the fusion data, the emotional intention of the home user is analyzed using the emotional intention analysis model.

[0140] The action grid refers to the division of a home environment into multiple virtual, interconnected areas. The transition probability matrix is ​​a probability model used to describe the movement of users between different grids. The emotional features are features extracted from the fused data that represent an individual's emotional state, such as facial expression features and body movement features. The emotional intention analysis model is a model used to analyze and understand the user's emotional state and the emotional tendencies they intend to express.

[0141] Optionally, the transition probability matrix of the household users can be constructed by a recurrent neural network, such as a long short-term memory network, a gated recurrent unit, etc.

[0142] Optionally, the home user's emotional intention analysis model can be constructed by machine learning techniques, such as support vector machines, random forests, neural networks, etc.

[0143] By analyzing the device parameter distribution of household appliances within a home environment, embodiments of the present invention can understand the usage frequency and energy consumption of different appliances at different time periods, thereby understanding the user's usage habits and preferences. The device parameter distribution refers to the distribution of operating parameters of various household appliances (such as lighting equipment, air conditioners, televisions, refrigerators, washing machines, etc.) at different times and spaces within a home environment.

[0144] Optionally, the device parameter distribution status of the home appliances in the home environment can be analyzed by big data analysis technology, such as data mining, machine learning algorithms, etc.

[0145] By determining the spatial preference matrix of the home user based on the high-frequency movement paths, embodiments of the present invention can identify the user's most frequently used areas and paths, thereby optimizing the home space in a personalized manner. The spatial preference matrix is ​​a mathematical model used to represent the degree of preference of the home user in different spatial areas.

[0146] As an embodiment of the present invention, determining the spatial preference matrix of the household user based on the high-frequency action path includes:

[0147] Encoding the high-frequency motion path to obtain an encoded trajectory;

[0148] Virtually constructing a spatial grid of the household user;

[0149] Mapping the encoded trajectory into a spatial grid to calculate the number of user visits and user stay time of the spatial grid;

[0150] Determine the spatial preference matrix of the home user according to the number of user visits and the user's stay time.

[0151] The coded trajectory refers to the conversion of a user's movement path within a home environment into a coded sequence in the form of numbers or symbols. The spatial grid divides the home environment into multiple small, regular areas to facilitate tracking and analysis of user activities within these areas. The number of user visits refers to the number of times a user enters or passes through a specific spatial grid area within a specific time period. The user dwell time refers to the duration of a user's stay within a specific spatial grid area.

[0152] Optionally, the spatial grid may be constructed by virtual reality technology.

[0153] By analyzing the user's psychological state profile based on the emotional intent, embodiments of the present invention can adjust the living environment based on the user's psychological state profile, providing the user with a more caring, healthy, and safe living environment. The psychological state profile is a multidimensional model that describes and represents the changes in an individual's psychological state over time.

[0154] Optionally, the psychological state map of the home user can be analyzed by a deep learning model, such as a convolutional neural network, a recurrent neural network, a long short-term memory network, etc.

[0155] By analyzing the functional distribution needs of home users based on the device parameter distribution, embodiments of the present invention can understand the user's usage preferences and needs for home devices in different time periods and scenarios. The functional distribution needs refer to the specific distribution of user needs for the functions provided by various home devices in a home environment across different spaces, times, and scenarios.

[0156] In this embodiment of the present invention, based on the spatial preference matrix, the psychological state map, and the functional distribution requirements, a preference fusion engine based on a preset intelligent AI analysis model is used to determine the personalized home preferences of the home user. The preference fusion engine can integrate this information to generate personalized home settings and automation scenarios, thereby providing a customized comfort experience. The preference fusion engine is a mathematical model used to integrate and process user preference data from different sources to generate a unified, personalized user preference configuration. Personalized home preferences refer to the user's personal preferences and needs for various home appliances, functions, environmental parameters, and spatial layout in their home living environment.

[0157] As an embodiment of the present invention, the method of determining the personalized home preferences of the home user based on the spatial preference matrix, the psychological state map, and the functional distribution requirements using a preference fusion engine of a preset intelligent AI analysis model includes:

[0158] defining a preference scoring function of the preference fusion engine;

[0159] Analyzing the spatial preference matrix, the mental state map, and the spatial preference weights, mental state weights, and functional requirement weights of the functional distribution requirements;

[0160] Determining the number of spatial regions, psychological dimensions, and functions of the spatial preference matrix, the psychological state map, and the functional distribution requirements;

[0161] Calculating the personalized home preference score of the home user using the preference score function according to the spatial preference weight, the psychological state weight, the functional requirement weight, the number of spatial areas, the psychological dimension, and the number of functions;

[0162] Based on the personalized home preference score, the personalized home preference of the home user is determined.

[0163] Among them, the preference score function refers to an algorithm used to comprehensively evaluate and quantify the user's overall preference level in the smart home environment. The spatial preference weight refers to the relative importance assigned to the spatial preference matrix information. The psychological state weight refers to the relative importance assigned to the psychological state map information. The functional requirement weight refers to the relative importance assigned to the functional requirement map information. The number of spatial areas refers to the number of independent areas into which the user's living space is divided when analyzing the user's spatial preferences. The psychological dimension refers to the different aspects or characteristics considered when analyzing the user's psychological state. The number of functions refers to the number of different functions provided by home devices. The personalized home preference score refers to the score calculated by the preference fusion engine in the smart home system, which is used to quantify the user's satisfaction with the current home environment or specific home functions.

[0164] Optionally, the preference score function of the preference fusion engine can be defined by a hypergraph attention mechanism.

[0165] S4. Based on the personalized home preferences, a personalized control scheme for the home devices of the home user is constructed, the real-time home information of the home user and the real-time environmental information of the home environment are collected, and the home environment is adaptively adjusted by using the personalized control scheme for the home devices in combination with the real-time home information and the real-time environmental information to obtain an adjusted home environment.

[0166] The embodiment of the application can analyze the preferences of the user at different times and in different situations, optimize the device parameter settings, and improve the comfort of the living environment by constructing the home device personalized control scheme of the home user based on the personalized home preferences. The home device personalized control scheme refers to a detailed and executable instruction or rule set for managing and adjusting the operating parameters of various devices in the smart home environment according to the personalized home preferences of the home user.

[0167] Optionally, the home device personalized control scheme of the home user can be constructed through a reward and punishment mechanism. For example, the home device parameters of the home user are determined through a reward and punishment mechanism based on the personalized home preferences, and the home device personalized control scheme of the home user is constructed based on the home device parameters.

[0168] The embodiment of the application can provide more personalized services by combining the real-time information of the user and the environmental information, understanding the current activities, positions, and needs of the user, and providing more personalized services. The real-time home information refers to data about the real-time state and behavior of the user at home. The real-time environmental information refers to real-time data about the current state of the user's living environment.

[0169] The embodiment of the application can significantly improve the comfort, safety, health, and energy efficiency of the living environment by using the home device personalized control scheme to combine the real-time home information and the real-time environmental information to adaptively adjust the home environment. The adjustment of the home environment refers to the automatic adjustment and control of various environmental parameters and device states in the home according to the real-time home information and the real-time environmental information, so as to create a more comfortable, safe, energy-saving, and personalized living environment.

[0170] As an embodiment of the application, the use of the home device personalized control scheme in combination with the real-time home information and the real-time environmental information to adaptively adjust the home environment to adjust the home environment includes:

[0171] Based on the real-time home information, the current user of the home environment is determined.

[0172] According to the personalized home preferences of the current user and the real-time environmental information, the personalized service needs of the current user are understood.

[0173] The demand changes of the personalized service needs are analyzed in real time, and the dynamic adjustment parameters of the corresponding home devices of the home environment are determined based on the demand changes.

[0174] Determining the control logic of the home device according to the personalized control solution of the home device;

[0175] According to the control logic and the dynamic adjustment parameters, the home environment is adaptively adjusted to obtain an adjusted home environment.

[0176] Among them, the current user refers to the family member who is in the home environment at a specific point in time and whose status, behavior or needs are being focused on and served by A1. The personalized service demand refers to the unique service that the user needs to achieve for the home environment (such as temperature, humidity, lighting, music, air, display content, etc.) at a specific moment, based on the current user's personal preferences, habits, physiological state, current activities and specific situation. The demand change refers to any changes and fluctuations in the personalized service needs of the current user within a specific time period. The dynamic adjustment parameter refers to the parameter of a specific device that needs to be dynamically controlled in order to achieve the environment adjustment goal. The control logic refers to a series of rules, algorithms and decision-making processes used to convert the environment adjustment goal into specific device adjustment parameters.

[0177] Optionally, the adjustment parameters of the home environment corresponding to the home devices can be determined by a reinforcement learning algorithm, such as Q-learning, deep Q network, etc.

[0178] S5. Collect the changes in facial expressions of the home user in the adjusted home environment in real time, analyze the comfort level of the home user with the current home environment based on the changes in facial expressions, and optimize the adjusted home environment based on the comfort level to obtain a target home environment.

[0179] By collecting real-time facial expressions of a user in the adjusted home environment, embodiments of the present invention can more accurately determine whether the current environmental adjustments meet the user's expectations and whether further adjustments are needed. The facial expressions refer to changes in the user's facial expressions in the home environment adjusted by the smart home system.

[0180] In this embodiment of the present invention, by analyzing the user's comfort level with their current home environment based on changes in facial expressions, facial recognition can be used to enable user control of their home environment without active interaction, thus achieving a more natural and implicit interaction method. The comfort level refers to the degree of physical and mental pleasure and relaxation a user experiences in their home environment.

[0181] As an embodiment of the present invention, analyzing the comfort level of the home user with the current home environment based on the expression change includes:

[0182] Determining a current activity of the home user, and analyzing the expression meaning of the expression change based on the current activity;

[0183] Calculating the expression intensity and expression duration of the expression change based on the expression meaning;

[0184] Classifying the expression changes according to the expression intensity and the expression duration to obtain multiple types of expressions, wherein the multiple types of expressions include: positive expressions, neutral expressions, and negative expressions;

[0185] Constructing a comfort mapping table of the home user to the current home environment according to the multiple types of expressions;

[0186] The expression change is mapped to the comfort mapping table to obtain the comfort level of the home user with the current home environment.

[0187] The current activity refers to the user's ongoing behavior in the smart home environment at a specific moment. The facial expression meaning refers to the emotional state, feelings, and intentions conveyed by the user's facial expression in the context of a specific activity. The expression intensity refers to the intensity or significance of the emotion expressed by the user's facial expression. The expression duration refers to the duration from the onset of a specific expression to its disappearance. The multiple categories of expressions refer to the classification of a user's facial expressions into multiple different categories based on factors such as the expression's meaning, intensity, and duration. Positive expressions refer to those within the user's facial expressions that express positive emotions. Such expressions generally reflect a good physical and mental state, such as comfort, pleasure, and satisfaction. Neutral expressions refer to those within the user's facial expressions that lack obvious positive or negative emotions. Such expressions generally reflect a relatively calm and peaceful state, without strong emotional fluctuations. Negative expressions refer to those within the user's facial expressions that express negative emotions. Such expressions generally reflect a negative physical and mental state, such as discomfort, displeasure, and dissatisfaction. The comfort mapping table refers to a table or database that associates various user expressions (positive, neutral, and negative) with the comfort parameters of the environment.

[0188] Optionally, the expression meaning of the expression change can be determined by natural language processing technology and emotion computing methods.

[0189] Optionally, the comfort mapping table can use VR technology to simulate different home environments to collect user comfort data in different situations, and construct a comfort mapping table for the home user based on the comfort data.

[0190] The embodiment of the present invention optimizes the adjusted home environment based on the comfort level to obtain a target home environment. Through accurate understanding of the user's comfort level, a highly personalized environment can be created to meet the unique needs of each user, thereby significantly improving the comfort and satisfaction of living.

[0191] Compared with the problems described in the background technology, the embodiment of the present invention can realize comprehensive perception of the home environment by collecting the environmental data of the home environment and the video image data and facial temperature data of the home users in the home environment based on the multi-sensor, thereby providing sufficient data sources for subsequent adaptive adjustment of the home scene; optionally, the embodiment of the present invention uses the data fusion module of the preset intelligent AI analysis model to fuse the environmental data, the behavioral data and the facial expression data, and the obtained fused data can more accurately analyze and meet the needs of the user, thereby improving the user experience; the embodiment of the present invention determines the personalized home preferences of the home users based on the spatial preference matrix, the psychological state map and the functional distribution requirements, and the preference fusion engine can integrate this information to generate a personalized home Home settings and automation scenes, thereby providing a customized comfort experience; the embodiment of the present invention utilizes the personalized control scheme of the home equipment in combination with the real-time home information and the real-time environmental information to adaptively adjust the home environment, and obtains that the adjusted home environment can significantly improve the comfort, safety, health and energy efficiency of the living environment, while achieving a higher level of automation and personalized services; the embodiment of the present invention can more accurately judge whether the current environmental adjustment has met the user's expectations and whether further adjustment is needed by collecting the changes in the expression of the home user in the adjusted home environment in real time; finally, the embodiment of the present invention optimizes the adjusted home environment based on the comfort level to obtain the target home environment. By accurately understanding the user's comfort level, it is possible to create a highly personalized environment to meet the unique needs of each user, thereby significantly improving the comfort and satisfaction of living. Therefore, the method and system for adaptive adjustment of home scenes driven by intelligent AI provided by the embodiment of the present invention can improve the accuracy of adaptive adjustment of home scenes.

[0192] Example 2:

[0193] like Figure 2 The figure shows a functional module diagram of a home scene adaptive adjustment system driven by intelligent AI in the present invention.

[0194] The intelligent AI-driven adaptive home scene adjustment system 200 described in the present invention can be installed in an electronic device. Depending on the functions implemented, the intelligent AI-driven adaptive home scene adjustment system can include a data acquisition module 201, a data fusion module 202, a personalized home preference analysis module 203, a home environment adjustment module 204, and a target home environment determination module 205. The modules described in the present invention can also be referred to as units, which refer to a series of computer program segments that can be executed by an electronic device processor and can perform fixed functions, and are stored in the memory of the electronic device.

[0195] In the embodiment of the present invention, the functions of each module / unit are as follows:

[0196] The data acquisition module 201 is used to configure multiple sensors in a home environment and collect environmental data of the home environment and video image data and facial temperature data of home users in the home environment based on the multiple sensors;

[0197] The data fusion module 202 is configured to extract behavioral features from the video image data, determine behavioral data of the household user based on the behavioral features, analyze facial expression data of the household user based on the facial temperature data and the video image data, and fuse the environmental data, the behavioral data, and the facial expression data using a data fusion module of a preset intelligent AI analysis model to obtain fused data;

[0198] The personalized home preference analysis module 203 is configured to determine the home user's high-frequency action paths and emotional intentions based on the fused data, analyze the device parameter distribution status of home devices in the home environment, determine the home user's spatial preference matrix based on the high-frequency action paths, analyze the home user's psychological state map based on the emotional intentions, analyze the home user's functional distribution requirements based on the device parameter distribution status, and determine the home user's personalized home preferences based on the spatial preference matrix, the psychological state map, and the functional distribution requirements using a preference fusion engine of a preset intelligent AI analysis model;

[0199] The home environment adjustment module 204 is configured to construct a home device parameter control scheme for the home user based on the personalized home preferences, collect real-time home information of the home user and real-time environmental information of the home environment, and adaptively adjust the home environment using the home device parameter control scheme in combination with the real-time home information and the real-time environmental information to obtain an adjusted home environment;

[0200] The target home environment determination module 205 is used to collect the changes in the home user's expression in the adjusted home environment in real time, analyze the home user's comfort level with the current home environment based on the changes in expression, and optimize the adjusted home environment based on the comfort level to obtain the target home environment.

[0201] In detail, the modules in the home scene adaptive adjustment system 200 driven by intelligent AI in the embodiment of the present invention are used in the same manner as above. Figure 1 The same technical means are used as the adaptive adjustment method for home scenes driven by intelligent AI described in , and can produce the same technical effects, so I will not go into details here.

[0202] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0203] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not limiting. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for adaptively adjusting home scenes driven by intelligent AI, characterized in that: The method comprises: Configuring multiple sensors within a home environment, and collecting environmental data of the home environment and video image data and facial temperature data of home users within the home environment based on the multiple sensors; extracting behavioral features from the video image data, determining behavioral data of the household user based on the behavioral features, analyzing facial expression data of the household user based on the facial temperature data and the video image data, and fusing the environmental data, the behavioral data, and the facial expression data using a data fusion module of a preset intelligent AI analysis model to obtain fused data; Based on the fused data, the high-frequency action paths and emotional intentions of the home user are determined, and the device parameter distribution status of the home devices in the home environment is analyzed. Based on the high-frequency action paths, the spatial preference matrix of the home user is determined. Based on the emotional intention, the psychological state map of the home user is analyzed. Based on the device parameter distribution status, the functional distribution requirements of the home user are analyzed. Based on the spatial preference matrix, the psychological state map, and the functional distribution requirements, the preference fusion engine of the preset intelligent AI analysis model is used to determine the personalized home preferences of the home user. Based on the personalized home preferences, a home device parameter control scheme is constructed for the home user, real-time home information of the home user and real-time environmental information of the home environment are collected, and the home environment is adaptively adjusted using the home device parameter control scheme in combination with the real-time home information and the real-time environmental information to obtain an adjusted home environment; The expression changes of the home user in the adjusted home environment are collected in real time, and based on the expression changes, the comfort level of the home user with the current home environment is analyzed. Based on the comfort level, the adjusted home environment is optimized to obtain a target home environment.

2. The method for adaptively adjusting home scenes driven by intelligent AI according to claim 1, wherein: The extracting the behavioral features of the video image data includes: Performing filtering and denoising processing on the video image data to obtain denoised video image data; Identifying sensitive areas in the denoised video image data, and performing pixelation desensitization processing on the denoised video image data according to the sensitive areas to obtain desensitized video image data; Identifying a target object of the desensitized video image data and generating a virtual target object of the target object; Determining a virtual character position of the virtual target object in consecutive frames; Based on the position of the virtual character, detecting key points of the body of the virtual target object; Calculating body motion parameters of the virtual target object based on the body key points; Extracting posture features and gesture features of the virtual target object; The behavior characteristics of the virtual target object are determined according to the body movement parameters, the posture characteristics, and the gesture characteristics.

3. The method for adaptively adjusting home scenes driven by intelligent AI according to claim 1, wherein: The analyzing the facial expression data of the household user based on the facial temperature data and the video image data includes: Identifying a target facial image of the video image data and generating a virtual facial image of the target facial image; Extracting facial feature points of the virtual face image; Extracting expression features of the virtual face image based on the facial feature points; determining a facial temperature distribution of the household user according to the facial temperature data; Identifying abnormal temperature areas on the face of the household user; determining a facial temperature feature of the household user according to the facial temperature distribution and the facial temperature abnormality area; Classifying the expressions of the household user according to the expression features and the facial temperature features to obtain basic emotions; Detecting slight changes in facial expressions of the household user, and determining facial expression data of the household user based on the slight changes in facial expressions and the basic emotions.

4. The method for adaptively adjusting home scenes driven by intelligent AI as claimed in claim 1, characterized in that: The data fusion module of the preset intelligent AI analysis model is used to fuse the environmental data, the behavioral data, and the facial expression data to obtain fused data, including: Analyzing the data association relationship among the environmental data, the behavioral data, and the facial expression data; extracting sensitive data of the environmental data, the behavioral data, and the facial expression data; Determining a sensitivity level of the sensitive data, and constructing a hierarchical dynamic encryption algorithm for the sensitive data based on the sensitivity level; Constructing a feature vector matrix of the environmental data, the behavioral data, and the facial expression data according to the data association relationship and the hierarchical dynamic encryption algorithm; Determining a feature weight matrix of the eigenvector matrix; According to the eigenvector matrix and the eigenweight matrix, a data fusion algorithm in the data fusion module is used to calculate a fusion eigenvector of the environmental data, the behavioral data, and the facial expression data, wherein the data fusion algorithm includes: Among them, R represents the fused feature vector, e represents the exponential function with e as the base, and T h represents the environmental eigenvector in the eigenvector matrix, q h represents the environmental feature weight in the feature weight matrix, T x represents the behavioral eigenvector in the eigenvector matrix, q x Represents the behavioral feature weight in the feature weight matrix, T b represents the expression eigenvector in the eigenvector matrix, T b represents the weight of the expression feature in the feature weight matrix, and C represents a constant; According to the fused feature vector, the data fusion module is utilized to output fused data of the environmental data, the behavioral data, and the facial expression data.

5. The method for adaptively adjusting home scenes driven by intelligent AI as claimed in claim 1, characterized in that: Determining the high-frequency action path and emotional intention of the household user based on the fused data includes: constructing an action grid of the household user based on the fused data; Constructing a transition probability matrix of the household user according to the action grid; According to the transition probability matrix, the high-frequency action path of the household user is calculated using the following formula: Among them, PathScore(G) represents the high-frequency action path, G represents the movement path of the home user, and n represents the number of grids in the action grid. Represents the transition probability matrix from the action grid g i Move to action grid g i+1 The probability of moving path G is , e represents the exponential function with base e, u represents the hyperparameter, and t(G) represents the total time taken to move path G; Extracting emotional features of the fused data, and constructing an emotional intention analysis model of the home user based on the emotional features; Based on the fusion data, the emotional intention of the home user is analyzed using the emotional intention analysis model.

6. The method for adaptively adjusting home scenes driven by intelligent AI according to claim 1, wherein: The determining of the spatial preference matrix of the household user based on the high-frequency action path includes: Encoding the high-frequency motion path to obtain an encoded trajectory; Virtually constructing a spatial grid of the household user; Mapping the encoded trajectory into a spatial grid to calculate the number of user visits and user stay time of the spatial grid; Determine the spatial preference matrix of the home user according to the number of user visits and the user's stay time.

7. The method for adaptively adjusting home scenes driven by intelligent AI as claimed in claim 1, characterized in that: The method of determining the personalized home preferences of the home user based on the spatial preference matrix, the psychological state map, and the functional distribution requirements and utilizing a preference fusion engine of a preset intelligent AI analysis model includes: defining a preference scoring function of the preference fusion engine; Analyzing the spatial preference matrix, the mental state map, and the spatial preference weights, mental state weights, and functional requirement weights of the functional distribution requirements; Determining the number of spatial regions, psychological dimensions, and functions of the spatial preference matrix, the psychological state map, and the functional distribution requirements; Calculating the personalized home preference score of the home user using the preference score function according to the spatial preference weight, the psychological state weight, the functional requirement weight, the number of spatial areas, the psychological dimension, and the number of functions; Based on the personalized home preference score, the personalized home preference of the home user is determined.

8. The method for adaptively adjusting home scenes driven by intelligent AI as claimed in claim 1, characterized in that: The method of utilizing the personalized control scheme for home devices in combination with the real-time home information and the real-time environment information to adaptively adjust the home environment to obtain an adjusted home environment includes: Determining a current user of the home environment based on the real-time home information; Understanding the personalized service needs of the current user based on the personalized home preferences corresponding to the current user and the real-time environmental information; Analyze changes in demand for personalized service needs in real time, and determine dynamic adjustment parameters of home devices corresponding to the home environment based on the changes in demand; Determining the control logic of the home device according to the personalized control solution of the home device; According to the control logic and the dynamic adjustment parameters, the home environment is adaptively adjusted to obtain an adjusted home environment.

9. The method for adaptively adjusting home scenes driven by intelligent AI as claimed in claim 1, characterized in that: Analyzing the comfort level of the home user with the current home environment based on the expression change includes: Determining a current activity of the home user, and analyzing the expression meaning of the expression change based on the current activity; Calculating the expression intensity and expression duration of the expression change based on the expression meaning; Classifying the expression changes according to the expression intensity and the expression duration to obtain multiple categories of expressions, wherein the multiple categories of expressions include: positive expressions, neutral expressions, and negative expressions; Constructing a comfort mapping table of the home user to the current home environment according to the multiple types of expressions; The expression change is mapped to the comfort mapping table to obtain the comfort level of the home user with the current home environment.

10. An intelligent AI-driven home scene adaptive adjustment system, characterized by: The system comprises: a data acquisition module, configured to configure multiple sensors within a home environment, and based on the multiple sensors, collect environmental data of the home environment and video image data and facial temperature data of home users within the home environment; a data fusion module, configured to extract behavioral features from the video image data, determine behavioral data of the household user based on the behavioral features, analyze facial expression data of the household user based on the facial temperature data and the video image data, and fuse the environmental data, the behavioral data, and the facial expression data using a data fusion module of a preset intelligent AI analysis model to obtain fused data; a personalized home preference analysis module for determining the home user's high-frequency action paths and emotional intentions based on the fused data, analyzing the device parameter distribution status of home devices in the home environment, determining the home user's spatial preference matrix based on the high-frequency action paths, analyzing the home user's psychological state map based on the emotional intentions, analyzing the home user's functional distribution requirements based on the device parameter distribution status, and determining the home user's personalized home preferences based on the spatial preference matrix, the psychological state map, and the functional distribution requirements using a preference fusion engine of a preset intelligent AI analysis model; a home environment adjustment module, configured to construct a home device parameter control scheme for the home user based on the personalized home preferences, collect real-time home information of the home user and real-time environmental information of the home environment, and adaptively adjust the home environment using the home device parameter control scheme in combination with the real-time home information and the real-time environmental information to obtain an adjusted home environment; The target home environment determination module is used to collect the changes in the home user's expression in the adjusted home environment in real time, analyze the home user's comfort with the current home environment based on the changes in expression, and optimize the adjusted home environment based on the comfort to obtain the target home environment.

Images