Terminal theme generation method and device, computer program product and electronic equipment

Abstract

The invention discloses a terminal theme generation method and device, a computer program product and electronic equipment. The method comprises the steps of obtaining multiple types of first physiological feature data of a target user using a target terminal in a first time period and multiple types of first environment feature data of an environment where the target user is located; analyzing the multiple types of first physiological feature data and the multiple types of first environment feature data by using a target fusion model, and determining a first emotion type of the target user in a first time period and a first environment type of an environment in which the target user is located, predicting a second emotion type of the target user in a second time period and a second environment type of the environment in which the target user is located; and determining a first terminal theme applied by the target terminal in the first time period according to the first emotion type, the first environment type, the second emotion type and the second environment type. The technical problem that the terminal theme for reflecting the current mood and the environment of the user cannot be intelligently generated in related technologies is solved.

Description

Technical Field

[0001] The present application relates to the field of terminal technology, and in particular to a method and device for generating a terminal theme, a computer program product, and an electronic device. Background Art

[0002] With the continuous development of terminal technology, various terminal devices have been gradually applied to people's lives and work. Usually, the application setting of terminal themes is often triggered by users actively, who choose to buy the theme they like in the theme store of the terminal system, while the terminal system is always in a passive acceptance state. Therefore, the setting method of terminal themes is rigid and inflexible, and cannot reflect the user's mood and current environment.

[0003] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0004] The embodiments of the present application provide a terminal theme generation method and device, a computer program product, and an electronic device to at least solve the technical problem that the related technology cannot intelligently generate a terminal theme that reflects the user's current mood and environment.

[0005] According to one aspect of an embodiment of the present application, a terminal theme generation method is provided, including: obtaining multiple categories of first physiological characteristic data of a target user using a target terminal in a first time period and multiple categories of first environmental characteristic data of the environment in which the user is located; using a pre-trained target fusion model to analyze the multiple categories of first physiological characteristic data and the multiple categories of first environmental characteristic data, respectively, to determine the first emotion type of the target user in the first time period and the first environmental type of the environment in which the user is located, and to predict the second emotion type of the target user in a second time period and the second environmental type of the environment in which the user is located, wherein the target fusion model includes: an emotion classification model composed of multiple first base models, an environmental classification model composed of multiple second base models, and each first base model is used to analyze a category of physiological characteristic data, and each second base model is used to analyze a category of environmental characteristic data, and the second time period is a time period after the first time period; determining the first terminal theme applied by the target terminal in the first time period based on the first emotion type, the first environmental type, the second emotion type, and the second environmental type.

[0006] Optionally, obtaining multiple categories of first physiological characteristic data of a target user using the target terminal within a first time period and multiple categories of first environmental characteristic data of the environment in which the target user is located includes: obtaining multiple categories of initial physiological characteristic data of the target user using the target terminal within the first time period and multiple categories of initial environmental characteristic data of the environment in which the target user is located, wherein the type of the initial physiological characteristic data includes at least one of the following: facial expression data, voice data, usage behavior data, and the type of the initial environmental characteristic data includes at least one of the following: location data, time, weather data, light intensity, and noise data; performing a first preprocessing operation on the multiple categories of initial physiological characteristic data to obtain multiple categories of first physiological characteristic data, and performing a second preprocessing operation on the multiple categories of initial environmental characteristic data to obtain multiple categories of first environmental characteristic data, wherein the first preprocessing operation includes at least one of the following: image enhancement processing, voice noise reduction, voice framing, and data standardization, and the second preprocessing operation includes at least one of the following: filtering processing, maximum and minimum standardization, outlier removal, and linear interpolation.

[0007] Optionally, the emotion classification model further includes: a first meta-learning model, and the environment classification model further includes a second meta-learning model, wherein the pre-trained target fusion model is used to analyze the plurality of first physiological characteristic data and the plurality of first environment characteristic data respectively to determine the first emotion type of the target user in the first time period and the first environment type of the environment in which the target user is located, including:

[0008] Each type of first physiological characteristic data is respectively input into the corresponding first base model in the emotion classification model for feature extraction to obtain the corresponding first physiological characteristic vector; the first physiological characteristic vector output by each first base model is input into the first meta-learning model in the emotion classification model for feature fusion to obtain the fused physiological characteristic vector, and the fused physiological characteristic vector is mapped to the predefined emotion state space to obtain the first emotion type of the target user in the first time period; each type of first environmental characteristic data is respectively input into the corresponding second base model in the environment classification model for feature extraction to obtain the corresponding first environmental characteristic vector; the first environmental characteristic vector output by each second base model is input into the second meta-learning model in the environment classification model for feature fusion to obtain the fused environmental characteristic vector, and the fused environmental characteristic vector is mapped to the predefined environmental state space to obtain the first environmental type of the environment in which the target user is located in the first time period.

[0009] Optionally, the target fusion model also includes: an emotion prediction model and an environment prediction model, wherein predicting the second emotion type of the target user in the second time period and the second environment type of the environment in which the target user is located includes: obtaining the third emotion type of the target user in the time period before the first time period and the third environment type of the environment in which the target user is located; using the emotion prediction model to analyze the first emotion type and the third emotion type, and predicting the second emotion type of the target user in the second time period, wherein the emotion prediction model is obtained by training using a first training sample set, and the first training sample set includes: the third emotion type of the target user in multiple consecutive third time periods and the third emotion type of the target user in the next time period of multiple consecutive third time periods, and the third time period is the time period before the first time period; using the environment prediction model to analyze the first environment type and the third environment type, and predicting the second environment type of the environment in which the target user is located in the second time period, wherein the environment prediction model is obtained by training using a second training sample set, and the second training sample set includes: the third environment type of the environment in which the target user is located in multiple consecutive third time periods and the third environment type of the environment in which the target user is located in the next time period of multiple consecutive third time periods.

[0010] Optionally, determining a first terminal theme applied by the target terminal in the first time period based on the first emotion type, the first environment type, the second emotion type and the second environment type includes: obtaining a third emotion type of the target user in a time period before the first time period and a third environment type of the environment in which the target user is located; constructing a first emotion portrait of the target user in the second time period based on the first emotion type, the second emotion type and the third emotion type, and constructing a first environment portrait of the target user in the second time period based on the first environment type, the second environment type and the third environment type; determining the first terminal theme applied by the target terminal in the first time period based on the first emotion portrait and the first environment portrait.

[0011] Optionally, determining a first terminal theme applied by the target terminal in the first time period based on the first emotion portrait and the first environment portrait includes: analyzing the first emotion portrait and the first environment portrait using a pre-trained target generation model to determine the first terminal theme applied by the target terminal in the first time period, wherein the first terminal theme includes at least one of: a background image, an application icon image.

[0012] Optionally, the training process of the target generation model includes: constructing an initial generation model; obtaining a third training sample set, wherein the third training sample set includes: multiple training samples consisting of second emotion portraits and second environment portraits of multiple users in different historical time periods, and the second terminal theme applied by each user in each historical time period as sample labels corresponding to the training samples; using the third training sample set to iteratively learn the initial generation model to obtain the target generation model.

[0013] According to another aspect of an embodiment of the present application, a terminal theme generating device is also provided, including: an acquisition module, used to acquire multiple categories of first physiological characteristic data of a target user using a target terminal in a first time period and multiple categories of first environmental characteristic data of the environment in which the user is located; an analysis module, used to use a pre-trained target fusion model to analyze the multiple categories of first physiological characteristic data and the multiple categories of first environmental characteristic data respectively, determine the first emotion type of the target user in the first time period and the first environmental type of the environment in which the user is located, and predict the second emotion type of the target user in the second time period and the second environmental type of the environment in which the user is located, wherein the target fusion model includes: an emotion classification model composed of multiple first base models, an environmental classification model composed of multiple second base models, and each first base model is used to analyze a category of physiological characteristic data, and each second base model is used to analyze a category of environmental characteristic data, and the second time period is a time period after the first time period; a determination module, used to determine the first terminal theme applied by the target terminal in the first time period based on the first emotion type, the first environmental type, the second emotion type, and the second environmental type.

[0014] According to another aspect of an embodiment of the present application, a computer program product is further provided, the computer program product comprising: a computer program, wherein when the computer program is executed by a processor, the above-mentioned terminal theme generating method is implemented.

[0015] According to another aspect of an embodiment of the present application, an electronic device is further provided, the electronic device comprising: a memory and a processor, wherein a computer program is stored in the memory, and the processor is configured to execute the above-mentioned terminal theme generation method through the computer program.

[0016] In the embodiment of the present application, the system analyzes the collected multiple categories of first physiological characteristic data and multiple categories of first environmental characteristic data through the pre-trained target fusion model, determines the first emotion type of the target user in the first time period and the first environmental type of the environment; and predicts the second emotion type of the target user in the second time period and the second environmental type of the environment by combining real-time data and historical data; finally, the first terminal theme applied by the target terminal in the first time period is determined according to the first environmental type, the second emotion type, the second emotion type and the second environmental type. The technical effect of dynamic adjustment and personalized generation of mobile phone themes according to user emotions and the environment is achieved, achieving the purpose of improving user experience, meeting user emotional expression and environmental adaptation needs, and solving the technical problem that related technologies cannot intelligently generate terminal themes that reflect the user's current mood and environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0018] Figure 1 It is a flowchart of an optional terminal theme generation method according to an embodiment of the present application;

[0019] Figure 2 is a schematic structural diagram of an optional terminal theme generating device according to an embodiment of the present application;

[0020] Figure 3 It is a schematic diagram of the structure of an optional electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0021] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.

[0022] It should be noted that the terms "first", "second", etc. in the specification, claims and drawings of the present application are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0023] Example 1

[0024] According to an embodiment of the present application, a terminal theme generation method is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0025] Figure 1 is a flow chart of a method for generating a terminal theme according to an embodiment of the present application, such as Figure 1 As shown, the method includes the following steps S102-S106, wherein:

[0026] Step S102, obtaining a plurality of first physiological characteristic data of a target user using a target terminal in a first time period and a plurality of first environmental characteristic data of the environment in which the target user is located.

[0027] In the technical solution provided in the above step S102, the target terminal refers to the smart device used by the target user, such as a smart phone, a smart watch, etc. The first time period refers to a period of time during which data collection and analysis are performed. Therefore, the system collects multiple categories of first physiological characteristic data of the target user in the first time period and multiple categories of first environmental characteristic data of the environment in which the target user is located by utilizing the multi-source sensors and cameras built into the target terminal, so as to subsequently fully and accurately understand the user's emotions and environment. Among them, the first physiological characteristic data can be understood as physiological indicator data directly related to the user's emotional state, which can reflect the user's current psychological state or emotional changes; the first environmental characteristic data is data that can describe the characteristics of the user's physical environment.

[0028] Step S104, using the pre-trained target fusion model to analyze the multiple categories of first physiological characteristic data and the multiple categories of first environmental characteristic data respectively, determine the first emotion type of the target user in the first time period and the first environmental type of the environment in which he is located, and predict the second emotion type of the target user in the second time period and the second environmental type of the environment in which he is located.

[0029] In the technical solution provided in the above step S104, the above target fusion model includes an emotion classification model composed of multiple first base models and an environment classification model composed of multiple second base models, and each first base model is used to analyze a type of physiological characteristic data, and each second base model is used to analyze a type of environmental characteristic data. Therefore, the system can first input the collected first physiological characteristic data and the first environmental characteristic data into the emotion classification model and the environmental classification model in the target fusion model for feature extraction, so as to respectively determine the first emotion type of the target user in the first time period and the first environmental type of the environment in which the target user is located. Furthermore, the model captures the law of feature change over time and predicts the second emotion type of the target user in the second time period and the second environmental type of the environment in which the target user is located based on the real-time data of the target user (i.e., multiple types of first physiological characteristic data of the target user in the first time period and multiple types of first environmental characteristic data of the environment in which the target user is located), wherein the above second time period is the time period after the first time period. In addition, the above first emotion type refers to the specific emotional state of the target user, which includes but is not limited to happiness, sadness, calmness, anxiety, etc.; and the first environmental type refers to the type of the environment in which the target user is located in the time period, such as work environment, natural environment, social environment, etc.

[0030] Step S106, determining a first terminal theme applied by the target terminal in the first time period according to the first emotion type, the first environment type, the second emotion type and the second environment type.

[0031] In the technical solution provided in the above step S106, the system determines a highly personalized and context-matching first terminal theme by analyzing the first emotion type of the target user in the first time period and the first environment type of the environment in which he is located (real-time data), and the second emotion type of the target user in the second time period and the second environment type of the environment in which he is located (predicted data), so as to better reflect the user's emotional state and the environment in which he is located, and enhance the user's usage experience and the human-computer interaction effect of the mobile phone.

[0032] Based on the scheme defined in the above steps S102 to S106, the system analyzes the collected multiple categories of first physiological characteristic data and multiple categories of first environmental characteristic data through the pre-trained target fusion model to determine the first emotion type of the target user in the first time period and the first environmental type of the environment in which he is located; and predicts the second emotion type of the target user in the second time period and the second environmental type of the environment in which he is located by combining real-time data and historical data; finally, the first terminal theme applied by the target terminal in the first time period is determined according to the first environmental type, the second emotion type, the second emotion type and the second environmental type. The technical effect of dynamic adjustment and personalized generation of mobile phone themes according to user emotions and the environment in which they are located is achieved, and the purpose of improving user experience, meeting user emotional expression and environmental adaptation needs is achieved, so as to solve the technical problem that the relevant technology cannot intelligently generate terminal themes that reflect the user's current mood and the environment in which they are located.

[0033] The specific implementation steps of the terminal theme generation method are described in detail below in conjunction with the specific implementation process.

[0034] As an optional implementation, in the technical solution provided in step S1021, the system may obtain the characteristic data of the target user according to the following method, including:

[0035] Step S1021, obtaining multiple types of initial physiological characteristic data of a target user using the target terminal in a first time period and multiple types of initial environmental characteristic data of the environment in which the target user is located.

[0036] The types of the above-mentioned initial physiological characteristic data include but are not limited to: facial expression data, voice data, usage behavior data, etc., and the types of initial environmental characteristic data include but are not limited to: location data, time, weather data, light intensity, noise data, etc. These data can be collected through sensors, cameras, microphones and other modules built into the target terminal.

[0037] For example, the system can collect initial physiological characteristic data through the following methods, including: using the built-in camera of the target terminal to capture the facial image of the target user to obtain the facial expression data of the target user; using the microphone of the target terminal to capture the voice signal of the target user when speaking to obtain the voice data of the target user; using the pressure sensor to monitor the target user's interactive behavior on the target terminal, such as application usage frequency, touch mode, input text content, etc., to obtain the target user's behavioral information.

[0038] In addition, the system can collect initial environmental characteristic data through the following methods, including: obtaining the location information of the target user through GPS technology; obtaining the real-time weather data (temperature, weather conditions, etc.) of the target user's location through the API interface; detecting the ambient light intensity and ambient noise level through the built-in sensors of the target terminal (such as light sensors, microphones).

[0039] Step S1022, performing a first preprocessing operation on the multiple categories of initial physiological characteristic data to obtain multiple categories of first physiological characteristic data, and performing a second preprocessing operation on the multiple categories of initial environmental characteristic data to obtain multiple categories of first environmental characteristic data.

[0040] Considering that the data collected by the sensor may contain noise, outliers, etc., in order to improve the data quality and ensure the accuracy of the model analysis results, the system can perform preprocessing operations on the initial information collected in the above step S1021. However, since the nature and collection methods of physiological characteristic data and environmental characteristic data are significantly different, different preprocessing operations can be used for these two types of data.

[0041] Specifically, since the initial physiological characteristic data contains a large amount of time series information and complex patterns, the first preprocessing operation includes but is not limited to: image enhancement processing, speech noise reduction, speech framing, data standardization, etc. For example, for the facial expression data in the initial physiological characteristic data, the system can use a facial detection algorithm to align and crop the image, focusing on the facial area; apply image enhancement techniques such as brightness and contrast adjustment, and denoising filtering to improve the accuracy of expression recognition. For the speech data in the initial physiological characteristic data, the system can perform noise reduction processing on the speech signal, such as using wavelet transform or spectral subtraction method; frame the continuous speech stream, and extract the features of each frame (such as Mel frequency cepstral coefficient MFCC, spectrogram) for subsequent sentiment analysis.

[0042] Specifically, since the initial environmental characteristic data have different physical dimensions and acquisition frequencies, the second preprocessing operation includes but is not limited to: filtering, maximum and minimum standardization, outlier detection and elimination based on the 3σ principle, linear interpolation, etc. For example, for the location data in the initial environmental characteristic data, the GPS (Global Positioning System) signal can be smoothed to reduce positioning errors; the geographic location can be encoded, such as using longitude and latitude or geocoding to convert it into a computable value. For the weather data in the initial environmental characteristic data, the format can be converted to extract key indicators (such as temperature, humidity, weather description); inconsistent weather descriptions can be uniformly encoded. For the light intensity in the initial environmental characteristic data, the maximum and minimum values ​​can be standardized to ensure the comparability of data under different lighting conditions; smoothing may also be required to reduce the interference of instantaneous light changes.

[0043] After obtaining the above-mentioned multiple first physiological characteristic data and multiple first environmental characteristic data, the system can input this information into the target fusion model, and use the emotion classification model and the environment classification model in the target fusion model for analysis. Among them, the emotion classification model adopts the Stacking model fusion technology, and extracts features through multiple basic learners (i.e., the first base model), wherein the multiple first base models in the emotion classification model include but are not limited to: Convolutional Neural Network (CNN), Recurrent Neural Network (RNN), Gradient Boosting Decision Tree (GBDT), etc.; and the environment classification model also adopts the Stacking model fusion technology, and extracts features through multiple basic learners (i.e., the second base model), wherein the multiple second base models in the environment classification model include but are not limited to: Gradient Boosting Decision Tree (GBDT), Convolutional Neural Network (CNN), Long Short-Term Memory (LSTM) network, etc. And the emotion classification model and the environment classification model base models are respectively used to process different types of feature data.

[0044] In addition, the above-mentioned emotion classification model also includes a first meta-learning model (also called a meta-learner), and the environment classification model also includes a second meta-learning model (also called a meta-learner).

[0045] Therefore, the system can call the target fusion model of the above architecture and analyze the feature data obtained in step S102 according to the following method to determine the first emotion type of the target user in the first time period and the first environment type of the environment in which the target user is located, including:

[0046] The first step: input each type of first physiological characteristic data into the corresponding first base model in the emotion classification model for feature extraction, and obtain the corresponding first emotion feature vector; input the first emotion feature vector output by each first base model into the first meta-learning model in the emotion classification model for feature fusion, and obtain the fused emotion feature vector; map the fused emotion feature vector to the predefined emotion state space, and obtain the first emotion type of the target user in the first time period.

[0047] For example, the embodiments of the present application can use a convolutional neural network to process the facial expression data of the target user, use a recurrent neural network to process voice data, and use a gradient boosting decision tree to process user behavior data, so as to obtain physiological feature vectors corresponding to each type of first physiological feature data respectively; then, the feature vectors output by these basic learners are input as new feature vectors into a logistic regression (LR) meta-learner (i.e., the first meta-learning model) for feature fusion to obtain the fused physiological feature vector, and the fused physiological feature vector is mapped to a predefined emotional state space to obtain the first emotion type of the target user in the first time period.

[0048] In the second step, each type of first environment feature data is input into the corresponding second base model in the environment classification model for feature extraction to obtain the corresponding first environment feature vector; the first environment feature vectors output by each second base model are input into the second meta-learning model in the environment classification model for feature fusion to obtain the fused environment feature vector, and the fused environment feature vector is mapped to the predefined environment state space to obtain the first environment type of the environment in which the target user is located in the first time period.

[0049] For example, the embodiments of the present application can use a gradient boosting decision tree to process environmental numerical data (such as temperature, light intensity, etc.), use a convolutional neural network to process scene image data, and use a long short-term memory network to process temporal data (position trajectory, etc.) to obtain environmental feature vectors corresponding to each type of first environmental feature data; then, the environmental feature vectors output by these basic learners are input as new feature vectors into a logistic regression (LR) meta-learner (i.e., the second meta-learning model) for feature fusion to obtain a fused environmental feature vector, and the fused environmental feature vector is mapped to a predefined environmental state space to obtain the first environmental type of the environment in which the target user is located in the first time period.

[0050] Furthermore, after the target user's first emotion type in the first time period and the first environment type in the environment, in order to more accurately characterize the target object's complete emotion change trend and the environment change trend, the embodiment of the present application can also predict the target user's second emotion type in the second time period and the second environment type in the environment through the emotion prediction model and the environment prediction model in the target fusion model, including:

[0051] Step 1: Obtain the third emotion type of the target user in the time period before the first time period and the third environment type of the environment in which the target user is located. In other words, the emotion type of the target user in the time period before the current time period is recorded as the third emotion type, and the environment type of the target user in the time period before the current time period is marked as the third environment type.

[0052] Step 2: Analyze the first emotion type and the third emotion type using the emotion prediction model to predict the second emotion type of the target user in the second time period.

[0053] The emotion prediction model is trained using the first training sample set, and the first training sample set includes: the third emotion type of the target user in a plurality of consecutive third time periods and the third emotion type of the target user in the next time period of the plurality of consecutive third time periods, and the third time period is the time period before the first time period. In other words, the emotion prediction model predicts the future trend of emotion changes by learning the evolution law of emotion.

[0054] Step 3: Analyze the first environment type and the third environment type using the environment prediction model to predict the second environment type of the target user in the second time period.

[0055] The above-mentioned environment prediction model is obtained by training using the second training sample set, and the second training sample set includes: the third environment type of the environment in which the target user is located in multiple consecutive third time periods and the third environment type of the environment in which the target user is located in the next time period of the multiple consecutive third time periods. In other words, the environment prediction model predicts the type of environment that the target user may be in in the future by analyzing the continuity of environmental changes.

[0056] It should be noted that the above-mentioned emotion prediction model and environment prediction model can be a time series analysis model, such as an autoregressive integrated moving average ARIMA model, a long short-term memory network (LSTM) or a gated recurrent unit (GRU), to model the emotion state sequence and the environment state sequence. These models can capture the law of emotion type and environment type changing over time.

[0057] As an optional implementation, in the technical solution provided in the above step S106, the system may determine the first terminal theme applied by the target terminal in the first time period according to the following method, including:

[0058] Step S1061, obtaining a third emotion type of the target user and a third environment type of the target user's environment in a time period before the first time period.

[0059] Specifically, the system can store the target user's emotion type and the environment type of the environment in each time period in the time series database for analysis. Therefore, the system can retrieve the target user's third emotion type and the third environment type of the environment in the previous time period of the first time period from the time series database.

[0060] Step S1062, constructing a first emotional portrait of the target user in the second time period based on the first emotion type, the second emotion type and the third emotion type, and constructing a first environmental portrait of the target user in the second time period based on the first environmental type, the second environmental type and the third environmental type.

[0061] The first emotional portrait is a combination of the current emotional state, the predicted future emotional state and the historical emotional trend to form a comprehensive emotional description that includes information such as emotional type, emotional intensity, and emotional fluctuation trend. The first environmental portrait is a combination of the current environmental state, the predicted future environmental state and the historical environmental change trend to form a comprehensive environmental description. It should be noted that the construction of the first emotional portrait can also take into account the user's personalized emotional threshold and preferences to provide a more personalized analysis.

[0062] Step S1063, determining a first terminal theme applied by the target terminal in the first time period according to the first emotion portrait and the first environment portrait.

[0063] Specifically, after obtaining the first emotion portrait and the first environment portrait, the system can use preset or self-learned theme generation rules to determine the first terminal theme that matches the user's emotional state and the environment in which he is located.

[0064] Optionally, in the technical solution provided in step S1063, the method can be implemented in the following manner: using a pre-trained target generation model to analyze the first emotional portrait and the first environmental portrait, and determining the first terminal theme applied by the target terminal in the first time period, wherein the first terminal theme includes but is not limited to: background images, application icon images, etc.

[0065] Specifically, the training process of the above target generation model is as follows:

[0066] First, an initial generation model is constructed, wherein the initial generation model may be a deep learning model.

[0067] Then, a third training sample set is obtained, wherein the third training sample set includes: a plurality of training samples consisting of second emotion portraits and second environment portraits of a plurality of users in different historical time periods, and a second terminal theme used by each user in each historical time period as a sample label corresponding to the training sample.

[0068] Finally, the third training sample set is used to iteratively learn the initial generation model to obtain the target generation model.

[0069] Therefore, the above target generation model can effectively analyze emotional portraits and environmental portraits to determine personalized mobile phone themes that meet the user's emotional and environmental needs.

[0070] In addition, the target generation model is used to analyze the first emotional portrait and the first environmental portrait, generate the corresponding first terminal theme, and feed back the first terminal theme to the target user. The target user can adjust the element layout, color matching, icon style, etc. according to their own needs to achieve the best balance between visual harmony and emotional expression, and feed back the adjusted first terminal theme to the system. The system can optimize the target generation model according to the first terminal theme, thereby forming a closed-loop optimization mechanism to continuously improve the accuracy of theme generation and user satisfaction.

[0071] Through the above-mentioned terminal theme generation method, the system generates personalized mobile phone themes that match the user's emotions and environmental status through real-time perception and prediction of the user's emotions and environment. First, the system can use the built-in sensors and cameras of the target terminal to collect multiple types of environmental feature data and multiple types of physiological feature data of the user; then, the system calls the target fusion model generated by the Stacking model fusion technology to analyze the collected multiple types of environmental feature data and multiple types of physiological feature data of the user, generate the user's emotion type and the environmental type of the environment in the current time period, and predict the user's emotion type and the environmental type of the environment in the future time period; finally, the first terminal theme applied by the target terminal in the first time period is determined based on the first emotion type, the first environmental type, the second emotion type and the second environmental type. Therefore, the present application scheme has the following technical advantages:

[0072] (1) Compared with the existing methods that usually rely on manual selection by the user or a fixed theme library, the embodiments of the present application can perceive the user's emotions and environmental changes in real time, and automatically generate themes that match the user's emotional state and the environment in which he is located, so that the mobile phone interface can intelligently respond to changes in the user's situation, improve the fit between the mobile phone interface and the user's emotional state, and provide a more personalized and emotional user experience.

[0073] (2) The embodiment of the present application adopts Stacking model fusion technology to combine multiple basic models, which improves the accuracy and robustness of environmental and emotional perception. Compared with the existing method of a single model, it can capture the user's multimodal information more comprehensively and accurately.

[0074] (3) Compared with pre-designed or periodically updated topics, the embodiments of the present application can achieve real-time generation, quickly respond to subtle changes in user emotions and environment, and ensure the immediacy and relevance of the topic.

[0075] (4) Existing methods cannot predict future emotional changes. However, the embodiments of the present application analyze historical data of user emotions, predict emotional change trends, and prepare or generate topics that may match future user emotions in advance, thereby improving the foresight and consistency of user experience.

[0076] (5) The existing method requires users to frequently switch themes manually, which is complex and time-consuming. However, the embodiment of the present application realizes the automatic real-time generation of themes, significantly reducing the complexity of the user's operation in theme selection and switching, and improving the convenience of use.

[0077] (6) The embodiments of the present application can generate diverse topic elements with limited resources, thus avoiding the waste and duplication of topic resources in the existing methods and improving resource utilization efficiency.

[0078] To sum up, the embodiments of the present application not only improve the intelligence and personalization level of theme generation, but also enhance the interactivity of the mobile phone interface with user emotions and environment, providing a more timely, accurate and emotional user experience, while reducing the complexity of user operations and achieving efficient use of resources.

[0079] Example 2

[0080] According to an embodiment of the present application, a terminal theme generation device for implementing the terminal theme generation method in embodiment 1 is also provided. Figure 2 As shown, the terminal theme generation device at least includes: an acquisition module 22, an analysis module 24 and a determination module 26, wherein:

[0081] An acquisition module 22, configured to acquire a plurality of first physiological characteristic data of a target user using a target terminal in a first time period and a plurality of first environmental characteristic data of the environment in which the target user is located;

[0082] The analysis module 24 is used to analyze the multiple categories of first physiological characteristic data and the multiple categories of first environmental characteristic data respectively by using the pre-trained target fusion model, determine the first emotion type of the target user in the first time period and the first environmental type of the environment in which the target user is located, and predict the second emotion type of the target user in the second time period and the second environmental type of the environment in which the target user is located, wherein the target fusion model includes: an emotion classification model composed of multiple first base models, an environmental classification model composed of multiple second base models, and each first base model is used to analyze a category of physiological characteristic data, and each second base model is used to analyze a category of environmental characteristic data, and the second time period is a time period after the first time period;

[0083] The determination module 26 is configured to determine a first terminal theme applied by the target terminal within a first time period according to the first emotion type, the first environment type, the second emotion type, and the second environment type.

[0084] It should be noted that each module in the terminal theme generation device in the embodiment of the present application corresponds one-to-one to each implementation step of the terminal theme generation method in Example 1. Since a detailed description has been given in Example 1, some details not reflected in this embodiment can be referred to Example 1 and will not be repeated here.

[0085] Example 3

[0086] According to an embodiment of the present application, a computer program product is also provided. The computer program product includes a computer program, wherein when the computer program is executed by a processor, the terminal theme generation method in Embodiment 1 is implemented.

[0087] According to an embodiment of the present application, a non-volatile storage medium is also provided, which includes a stored computer program, wherein the device where the non-volatile storage medium is located executes the terminal theme generation method in Example 1 by running the computer program.

[0088] According to an embodiment of the present application, a processor is further provided, which is used to run a computer program, wherein the terminal theme generation method in Example 1 is executed when the computer program is running.

[0089] According to an embodiment of the present application, an electronic device is also provided, which includes: a memory and a processor, wherein a computer program is stored in the memory, and the processor is configured to execute the terminal theme generation method in Example 1 through the computer program.

[0090] Optionally, when the computer program is running, the following steps are executed: obtaining multiple categories of first physiological characteristic data of a target user using the target terminal in a first time period and multiple categories of first environmental characteristic data of the environment in which the user is located; using a pre-trained target fusion model to analyze the multiple categories of first physiological characteristic data and the multiple categories of first environmental characteristic data, respectively, to determine the first emotion type of the target user in the first time period and the first environmental type of the environment in which the user is located, and to predict the second emotion type of the target user in a second time period and the second environmental type of the environment in which the user is located, wherein the target fusion model includes: an emotion classification model composed of multiple first base models, an environmental classification model composed of multiple second base models, and each first base model is used to analyze a category of physiological characteristic data, and each second base model is used to analyze a category of environmental characteristic data, and the second time period is a time period after the first time period; determining the first terminal theme applied by the target terminal in the first time period based on the first emotion type, the first environmental type, the second emotion type, and the second environmental type.

[0091] As an optional implementation, the electronic device may be in the form of a mobile terminal, a computer terminal or a similar computing device. Figure 3 FIG. 1 shows a hardware structure block diagram of an electronic device for implementing a terminal theme generation method. Figure 3 As shown, the electronic device 30 may include one or more (302a, 302b, ..., 302n are used to illustrate) processors 302 (the processor 302 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA), a memory 304 for storing data, and a transmission device 306 for communication functions. In addition, it may also include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of the BUS bus), a network interface, a power supply and / or a camera. It can be understood by those skilled in the art that Figure 3 The structure shown is only for illustration and does not limit the structure of the above electronic device. Figure 3 More or fewer components as shown, or with Figure 3 Different configurations shown.

[0092] It should be noted that the one or more processors 302 and / or other data processing circuits described above may generally be referred to herein as "data processing circuits". The data processing circuits may be embodied in whole or in part as software, hardware, firmware, or any other combination thereof. In addition, the data processing circuit may be a single independent processing module, or may be incorporated in whole or in part into any of the other components in the electronic device 30. As described in the embodiments of the present application, the data processing circuit acts as a processor control (e.g., selection of a variable resistor terminal path connected to an interface).

[0093] The memory 304 can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the terminal theme generation method in the embodiment of the present application. The processor 302 executes various functional applications and data processing by running the software programs and modules stored in the memory 304, that is, realizing the vulnerability detection method of the above-mentioned application. The memory 304 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 304 may further include a memory remotely arranged relative to the processor 302, and these remote memories may be connected to the electronic device 30 via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0094] The transmission device 306 is used to receive or send data via a network. The specific example of the above network may include a wireless network provided by a communication provider of the electronic device 30. In one example, the transmission device 306 includes a network adapter (Network Interface Controller, NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 306 can be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0095] The display may be, for example, a touch screen liquid crystal display (LCD) that enables a user to interact with a user interface of the electronic device 30 .

[0096] The serial numbers of the above embodiments are only for description and do not represent the advantages or disadvantages of the embodiments.

[0097] In the above embodiments of the present application, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, please refer to the relevant description of other embodiments.

[0098] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of units can be a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0099] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed over multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0100] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0101] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art or all or part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, server or network device, etc.) to perform all or part of the steps of each embodiment method of the present application. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, disk or optical disk, etc., which can store program code.

[0102] The above are only preferred implementations of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A method for generating a terminal theme, characterized in that: include: Acquire multiple categories of first physiological characteristic data of a target user using a target terminal in a first time period and multiple categories of first environmental characteristic data of the environment in which the target user is located; The pre-trained target fusion model is used to analyze the multiple categories of first physiological characteristic data and the multiple categories of first environmental characteristic data respectively, determine the first emotion type of the target user in a first time period and the first environmental type of the environment in which the target user is located, and predict the second emotion type and the second environmental type of the environment in a second time period, wherein the target fusion model includes: an emotion classification model composed of a plurality of first base models, an environmental classification model composed of a plurality of second base models, and each of the first base models is used to analyze a category of physiological characteristic data, and each of the second base models is used to analyze a category of environmental characteristic data, and the second time period is a time period after the first time period; A first terminal theme applied by the target terminal in a first time period is determined according to the first emotion type, the first environment type, the second emotion type, and the second environment type.

2. The method according to claim 1, characterized in that Acquiring multiple types of first physiological characteristic data of a target user using a target terminal in a first time period and multiple types of first environmental characteristic data of the environment in which the target user is located, including: Acquire multiple types of initial physiological characteristic data of the target user using the target terminal in a first time period and multiple types of initial environmental characteristic data of the environment in which the target user is located, wherein the type of the initial physiological characteristic data includes at least one of the following: facial expression data, voice data, and usage behavior data; the type of the initial environmental characteristic data includes at least one of the following: location data, time, weather data, light intensity, and noise data; A first preprocessing operation is performed on the multiple categories of initial physiological characteristic data to obtain the multiple categories of first physiological characteristic data, and a second preprocessing operation is performed on the multiple categories of initial environmental characteristic data to obtain the multiple categories of first environmental characteristic data, wherein the first preprocessing operation includes at least one of the following: image enhancement processing, speech noise reduction, speech framing, and data standardization, and the second preprocessing operation includes at least one of the following: filtering processing, maximum and minimum standardization, outlier elimination, and linear interpolation.

3. The method according to claim 1, characterized in that The emotion classification model further includes: a first meta-learning model, and the environment classification model further includes a second meta-learning model, wherein the plurality of first physiological characteristic data and the plurality of first environment characteristic data are analyzed respectively by using a pre-trained target fusion model to determine the first emotion type of the target user in the first time period and the first environment type of the environment in which the target user is located, including: Input each type of the first physiological characteristic data into the corresponding first base model in the emotion classification model for feature extraction to obtain the corresponding first physiological characteristic vector; input the first physiological characteristic vector output by each first base model into the first meta-learning model in the emotion classification model for feature fusion to obtain the fused physiological characteristic vector, and map the fused physiological characteristic vector into a predefined emotional state space to obtain the first emotion type of the target user in the first time period; Input each type of the first environmental feature data into the corresponding second base model in the environmental classification model for feature extraction to obtain the corresponding first environmental feature vector; input the first environmental feature vector output by each second base model into the second meta-learning model in the environmental classification model for feature fusion to obtain a fused environmental feature vector, and map the fused environmental feature vector to a predefined environmental state space to obtain the first environmental type of the environment in which the target user is located in the first time period.

4. The method according to claim 3, characterized in that The target fusion model also includes: an emotion prediction model and an environment prediction model, wherein predicting the second emotion type of the target user and the second environment type of the environment in the second time period includes: Acquire a third emotion type and a third environment type of the target user in a time period before the first time period; Analyzing the first emotion type and the third emotion type using the emotion prediction model to predict the second emotion type of the target user in a second time period, wherein the emotion prediction model is obtained by training using a first training sample set, and the first training sample set includes: the third emotion type of the target user in a plurality of consecutive third time periods and the third emotion type of the target user in a next time period of the plurality of consecutive third time periods, and the third time period is a time period before the first time period; The first environment type and the third environment type are analyzed using the environment prediction model to predict the second environment type of the target user's environment in a second time period, wherein the environment prediction model is trained using a second training sample set, and the second training sample set includes: the third environment type of the target user's environment in multiple consecutive third time periods and the third environment type of the target user's environment in the next time period of the multiple consecutive third time periods.

5. The method according to claim 1, characterized in that: Determining a first terminal theme applied by the target terminal in a first time period according to the first emotion type, the first environment type, the second emotion type, and the second environment type includes: Acquire a third emotion type and a third environment type of the target user in a time period before the first time period; Constructing a first emotion portrait of the target user in a second time period according to the first emotion type, the second emotion type, and the third emotion type, and constructing a first environment portrait of the target user in the second time period according to the first environment type, the second environment type, and the third environment type; A first terminal theme applied by the target terminal in a first time period is determined based on the first emotion portrait and the first environment portrait.

6. The method according to claim 5, characterized in that Determining a first terminal theme applied by the target terminal in a first time period according to the first emotion portrait and the first environment portrait includes: The first emotion portrait and the first environment portrait are analyzed using a pre-trained target generation model to determine a first terminal theme applied by the target terminal in a first time period, wherein the first terminal theme includes at least one of the following: a background image, an application icon image.

7. The method according to claim 6, characterized in that The training process of the target generation model includes: Build an initial generative model; Obtaining a third training sample set, wherein the third training sample set includes: a plurality of training samples consisting of second emotion portraits and second environment portraits of a plurality of users in different historical time periods, and a second terminal theme used by each user in each historical time period as a sample label corresponding to the training samples; The initial generation model is iteratively learned using the third training sample set to obtain the target generation model.

8. A terminal theme generating device, characterized in that: include: An acquisition module, used to acquire multiple types of first physiological characteristic data of a target user using a target terminal in a first time period and multiple types of first environmental characteristic data of the environment in which the target user is located; The analysis module is used to analyze the multiple categories of first physiological characteristic data and the multiple categories of first environmental characteristic data respectively by using the pre-trained target fusion model, determine the first emotion type of the target user in the first time period and the first environmental type of the environment in which the target user is located, and predict the second emotion type of the target user in the second time period and the second environmental type of the environment in which the target user is located, wherein the target fusion model includes: An emotion classification model composed of a plurality of first base models, and an environment classification model composed of a plurality of second base models, wherein each of the first base models is used to analyze a type of physiological characteristic data, each of the second base models is used to analyze a type of environmental characteristic data, and the second time period is a time period after the first time period; A determination module is used to determine a first terminal theme applied by the target terminal within a first time period according to the first emotion type, the first environment type, the second emotion type and the second environment type.

9. A computer program product, characterized in that include: A computer program, wherein when the computer program is executed by a processor, the terminal theme generation method described in any one of claims 1 to 7 is implemented.

10. An electronic device, characterized in that: include: A memory and a processor, wherein the memory stores a computer program, and the processor is configured to execute the terminal theme generating method according to any one of claims 1 to 7 through the computer program.

Images