An emotion computing method and system based on multi-site skin physiological response

Abstract

The application provides an emotion calculation method and system based on multi-site skin physiological response, comprising: acquiring multi-site skin electric data corresponding to different emotional states; decomposing and extracting features of the multi-site skin electric data to generate a training data set and a test data set; training a preset prediction model through the training data set and the test data set to obtain a final prediction model; analyzing real-time skin electric data through the final prediction model to acquire a current emotional state; wherein the final prediction model is obtained by training the preset prediction model through the training data set and testing the trained prediction model through the test data. The application solves the problem of inaccurate human emotion perception prediction in the prior art.

Description

Technical Field

[0001] This invention relates to the field of emotion perception technology, and in particular to an emotion computing method and system based on multi-site skin physiological responses. Background Technology

[0002] Affective computing technology refers to the technology of using computers to identify, understand, and process information collected from emotions. As an interdisciplinary technology, research on affective computing involves multiple fields such as computer science, psychology, and cognitive science. Among the various components of emotion, physiological signals are difficult to fake and can be conveniently and continuously acquired without being noticed. Therefore, in recent years, affective computing methods have increasingly focused on collecting and computing various peripheral physiological signals, such as skin conductance and pulse wave signals.

[0003] However, to date, emotion computing technologies based on electrodermatology (EDS) primarily limit their measurement sites to single body parts, such as fingers or wrists, rarely considering the spatial distribution of bodily responses. This limitation first affects the accuracy of emotion computing; single data sources are more susceptible to interference from movement and environmental factors, making it difficult to discover stable correlations between EDS and emotions, and frequently resulting in contradictory findings. Furthermore, emotional bodily responses are systemic and difficult to record completely from a single site. For example, there may be left-right asymmetry in EDS under different emotional states, and body activation maps drawn based on subjective feelings also reveal overall differences across different emotions. Classical emotion theory posits that the overall perception of bodily responses influences subjective emotional feelings. However, current methods of measuring emotion at multiple single points make it difficult to accurately perceive human emotional states based on multi-point measurements. Summary of the Invention

[0004] This invention provides an emotion computing method and system based on multi-site skin physiological responses to solve the problem of inaccurate prediction of human emotion perception in existing technologies.

[0005] This invention provides an emotion computing method based on multi-site skin physiological responses, comprising:

[0006] Acquire multi-site electrodermal data corresponding to different emotional states;

[0007] The multi-site electrodermal data are decomposed and feature extracted to generate training and testing datasets;

[0008] The preset prediction model is trained using the training dataset and the test dataset to obtain the final prediction model;

[0009] The final prediction model is used to analyze real-time skin conductance data to obtain the current emotional state.

[0010] The final prediction model is obtained by training a preset prediction model using the training dataset and testing the trained prediction model using the test data.

[0011] The present invention provides an emotion computing method based on multi-site skin physiological response, which uses a multi-site skin electrosensing device to adhere to the human skin at different locations;

[0012] The skin electrosensing device measures the resistance between two electrodes by discharging electrodes and generates a resistance value.

[0013] Based on the resistance values, multi-site electrodermal data corresponding to different emotional states are generated.

[0014] According to the present invention, an emotion computing method based on multi-site skin physiological responses decomposes and extracts features from the multi-site electrodermatology data to generate training and testing datasets, specifically including:

[0015] The multi-site electrodermal data were decomposed into fast-varying and slow-varying electrodermal responses using a continuous decomposition analysis method.

[0016] Feature extraction of the multi-point electrodermal data includes time-domain feature extraction and frequency-domain feature extraction;

[0017] The data after feature extraction is divided equally to generate training and testing datasets.

[0018] According to the present invention, an emotion computing method based on multi-site skin physiological response is provided, which performs feature extraction on the multi-site electrodermatology data, including time-domain feature extraction and frequency-domain feature extraction, specifically including:

[0019] The time-domain features include standard deviation, mean, root mean square, and higher-order statistical features. For short-term event-related response analysis, waveform features of physiological changes are added.

[0020] The frequency domain features include peak frequency and harmonics. Higher-order continuous signal features can be extracted by performing discrete wavelet transformation and empirical mode decomposition on the signal.

[0021] According to the present invention, an emotion computing method based on multi-site skin physiological response is provided, which trains a preset prediction model using the training dataset and the test dataset to obtain a final prediction model, specifically including:

[0022] For categorized emotion tags, including positive or negative emotions and arousal levels, construct a classification prediction model;

[0023] A regression prediction model is constructed based on continuous emotional ratings.

[0024] The classification and regression prediction models were cross-validated. For data within a single subject, the data were evenly divided into a training set and a test set. The classification and regression models were trained on the training set, and the final prediction model was selected based on the accuracy of the prediction results on the test set.

[0025] According to the present invention, an emotion computing method based on multi-site skin physiological response analyzes real-time electroskin data through the final prediction model to obtain the current emotional state, specifically including:

[0026] The real-time acquired skin conductance data is sent to the final prediction model;

[0027] The final prediction model is used for analysis and matching to output the predicted emotional state.

[0028] This invention also provides an emotion computing system based on multi-site skin physiological responses, the system comprising:

[0029] The data acquisition module is used to acquire multi-site electroskin data corresponding to different emotional states;

[0030] The data processing module is used to decompose and extract features from the multi-site electrodermatology data to generate training and test datasets.

[0031] The model building module is used to train a preset prediction model using the training dataset and the test dataset to obtain the final prediction model.

[0032] The prediction module is used to analyze real-time skin conductance data through the final prediction model to obtain the current emotional state;

[0033] The final prediction model is obtained by training a preset prediction model using the training dataset and testing the trained prediction model using the test data.

[0034] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the emotion computing method based on multi-site skin physiological response as described above.

[0035] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the emotion computing method based on multi-site skin physiological response as described above.

[0036] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the emotion computing method based on multi-site skin physiological response as described above.

[0037] This invention provides an emotion computing method and system based on multi-site skin physiological responses. By using sensors capable of measuring skin conductance responses at multiple sites, it induces emotional states in participants using specific emotional materials and records the current changes in skin conductance at these sites, establishing an automated emotion recognition method. Compared to traditional single-site skin conductance measurements, the method proposed in this invention considers the body distribution of emotion-induced skin conductance responses, which is beneficial for stably and completely discovering the correlation between emotions and skin conductance responses, improving the accuracy of emotion recognition. Furthermore, the wearable method itself has a low workload, which helps promote the widespread adoption of everyday wearable emotion measurement and has significant application value for the development of emotion computing and emotional intelligence. Attached Figure Description

[0038] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0039] Figure 1 This is one of the flowcharts of an emotion computing method based on multi-site skin physiological response provided by the present invention;

[0040] Figure 2 This is the second flowchart of an emotion computing method based on multi-site skin physiological response provided by the present invention;

[0041] Figure 3 This is the third flowchart of an emotion computing method based on multi-site skin physiological response provided by the present invention;

[0042] Figure 4 This is the fourth flowchart of an emotion computing method based on multi-site skin physiological response provided by the present invention;

[0043] Figure 5 This is the fifth flowchart of an emotion computing method based on multi-site skin physiological response provided by the present invention;

[0044] Figure 6 This is a schematic diagram of the module connections of an emotion computing system based on multi-site skin physiological responses provided by the present invention;

[0045] Figure 7This is a schematic diagram of the multi-site electrodermal measurement sites provided by the present invention;

[0046] Figure 8 This is a schematic diagram of the structure of the electronic device provided by the present invention.

[0047] Figure label:

[0048] 110: Data acquisition module; 120: Data processing module; 130: Model building module; 140: Prediction module;

[0049] 810: Processor; 820: Communication interface; 830: Memory; 840: Communication bus. Detailed Implementation

[0050] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0051] The following is combined with Figures 1-5 This invention describes an emotion computing method based on multi-site skin physiological responses, comprising:

[0052] S100: Obtain multi-site electrodermal data corresponding to different emotional states;

[0053] S200: Decompose and extract features from the multi-site electrodermal data to generate training and test datasets;

[0054] S300. The preset prediction model is trained using the training dataset and the test dataset to obtain the final prediction model;

[0055] S400. Analyze the real-time skin conductance data using the final prediction model to obtain the current emotional state;

[0056] The final prediction model is obtained by training a preset prediction model using the training dataset and testing the trained prediction model using the test data.

[0057] This invention utilizes a multi-channel electrodermal (EDS) sensor to record EDS signals from multiple body parts under different emotional states. Signal decomposition, feature extraction, and prediction models are then performed to construct a wearable, multi-site automated emotion computing method. This addresses the shortcomings of existing physiological measurement methods, which primarily rely on single body parts and struggle to reveal the full picture of the relationship between emotions and bodily responses, effectively improving the accuracy of emotion recognition. Furthermore, by demonstrating the convenience and low-impact nature of wearable physiological measurement, this invention also possesses significant application value, contributing to the widespread adoption of peripheral physiological measurement and emotion computing technologies.

[0058] Acquire multi-site electrodermal data corresponding to different emotional states, specifically including:

[0059] S101, The skin is attached to different parts of the human skin through a multi-point electrodermal sensing device;

[0060] S102. The skin electrosensing device measures the resistance between two electrodes by discharging electrodes and generates a resistance value.

[0061] S103. Generate multi-site electrodermal data corresponding to different emotional states based on the resistance values.

[0062] The multi-site electrodermal (EDS) sensing device in this invention forms the basis of multi-site emotion computing. A common method for measuring EDS involves applying a small voltage to two electrodes on the skin surface and measuring the resistance between them. Since skin resistance is typically above 100 kΩ, and can reach tens of megohms, the EDS measurement device should have a wide measurement range to cover this range. To completely separate the EDS response waveform, a general sampling rate of approximately 200–400 Hz is recommended, with a minimum of 20 Hz. Simultaneously, multi-site measurements require multiple EDS sensors to operate simultaneously, and the time accuracy of the recorded data must also meet requirements; that is, for measurements of approximately 3 hours, the time difference between sensors should not exceed 0.5 seconds. To achieve this, this method requires the EDS acquisition hardware or software system to have a high-precision and stable timing function, meaning that each sampling point's data includes information about the actual acquisition time. Furthermore, the hardware and software system must have time alignment capabilities, i.e., synchronization with other experimental devices such as stimulus presentation systems.

[0063] Considering that experimenters need to understand the device's operating status and data acquisition, the sensing device should have data transmission capabilities to wirelessly transmit data to the receiving end and provide real-time visualization of the data. Simultaneously, considering potential packet loss and latency in wireless transmission, the sensor should also have local memory card writing capabilities to accurately and completely preserve high-sampling-rate electrodermal data.

[0064] In actual data acquisition, the performance of the sensor used is determined based on the parameters required for multi-site skin resistance / conductivity measurement and analysis. Specific performance considerations include: the measurement range (scale) of skin resistance / conductivity, sampling rate, and data transmission or storage method. If wired transmission is used, device size, wearing comfort, and connection stability must be considered. If wireless devices are used, transmission latency and battery life must be considered. These performance parameters should encompass all the requirements of the measurement; for example, the measurement range should cover the possible range of the measured object.

[0065] Multi-site electrodermal (EDS) sensors should be matched and calibrated with standardized physiological polysaccharides. This can be done by fixing two sets of devices to close locations and comparing the measurement results of the two sets of devices during a specific task, such as deep breathing that induces changes in EDS. If the measurement results and trends of the two sets of devices are similar, the next step of data acquisition can be supported. If there are significant differences between the two sets of measurements, the internal components of the devices should be checked and adjusted to ensure the quality and validity of the acquired data.

[0066] During the data collection phase, each participant wore the same multi-site kinetic sensor for data collection. The data sampling rate was no less than 20Hz. Sensor measurement sites included, but were not limited to: both wrists, both ankles, back of the neck, back, diaphragm, and both sides of the waist. Specific sites can be found by referring to [reference needed]. Figure 7 ,exist Figure 7 The numbers 1-9 represent different sensor measurement sites. Increasing the number of measurement sites can improve the accuracy of emotion detection and recognition.

[0067] Each participant evoked emotions using specific standardized emotional evoked materials. Each material had its own labeled emotional attributes, such as valence and arousal. After each task, participants reported their current emotional state via a subjective questionnaire. Rest periods were set between materials, and the rest period between groups was no less than 30 seconds. The emotional labels of the emotional materials and the subjective emotional state reported after each task were used as dependent variables.

[0068] The multi-site electrodermal data are decomposed and feature extracted to generate training and testing datasets, specifically including:

[0069] S201. The multi-site electrodermal data is decomposed into fast-changing electrodermal responses and slow-changing electrodermal responses using a continuous decomposition analysis method.

[0070] S202, Feature extraction of the multi-point electrodermal data includes time-domain feature extraction and frequency-domain feature extraction;

[0071] S203. Divide the data after feature extraction into equal parts to generate training and test datasets.

[0072] In the analysis and calculation of electrodermal signals (EDS), the EDS is typically decomposed into two components using Continuous Decomposition Analysis (CDA): a rapidly changing EDS response and a slowly changing EDS response. The former is usually triggered by a specific stimulus event and changes on a timescale of seconds, while the latter changes more gradually. Features frequently extracted from EDS signals include time-domain and frequency-domain features. These features are used as inputs to emotion computing models. Time-domain features include basic characteristics such as standard deviation, mean, and root mean square (RMS), as well as higher-order statistical (HOC) features such as skewness and kurtosis. Furthermore, for short-term event-related response analysis, waveform characteristics of physiological changes are also considered, such as SCR amplitude, mean rise time, or sum of SCR areas. Frequency-domain features include peak frequency and harmonics. Additionally, performing discrete wavelet transforms and empirical mode decomposition methods on the signal can help extract higher-order continuous signal features.

[0073] The preset prediction model is trained using the training dataset and the test dataset to obtain the final prediction model, specifically including:

[0074] S301. Construct a classification prediction model for categorized emotion labels, including positive / negative emotions and arousal levels.

[0075] S302. Construct a regression prediction model based on continuous emotional ratings;

[0076] S303. The classification prediction model and regression prediction model are cross-validated. For the data within a single subject, the data is evenly divided into a training set and a test set. The classification and regression models are trained on the training set, and the final prediction model is selected based on the accuracy of the prediction results on the test set.

[0077] This invention constructs a classification prediction model for categorized emotion labels, such as positive / negative emotions and arousal levels; and a regression prediction model for continuous emotion ratings. The model employs cross-validation, uniformly dividing the data within a single subject into training and test sets. The classification and regression models are trained on the training set, and the accuracy of the predictions on the test set is used to evaluate the model's performance, allowing for adjustments to improve its effectiveness.

[0078] The final prediction model analyzes real-time skin conductance data to obtain the current emotional state, specifically including:

[0079] S401. Send the real-time acquired skin conductance data to the final prediction model;

[0080] S402. Analyze and match the data using the final prediction model to output the predicted emotional state.

[0081] For a new participant, by wearing a wearable electrodermal sensor, the electrodermal response of multiple parts of their body under a specific emotional state is recorded. Using the same electrodermal processing and feature extraction methods as the training process described above, the current emotional state of the participant is finally obtained as the predicted output.

[0082] This invention provides an emotion computing method based on multi-site skin physiological responses. By using sensors capable of measuring skin conductance responses at multiple sites, it induces participants' emotional states using specific emotional material and records the current changes in skin conductance at these sites, establishing an automated emotion recognition method. Compared to traditional single-site skin conductance measurements, the proposed method considers the body distribution of emotion-induced skin conductance responses, which is beneficial for stably and completely discovering the correlation between emotions and skin conductance responses, improving the accuracy of emotion recognition. Furthermore, the wearable method itself has a low workload, which helps promote the widespread adoption of everyday wearable emotion measurement and has significant application value for the development of emotion computing and emotional intelligence.

[0083] refer to Figure 6 The present invention also discloses an emotion computing system based on multi-site skin physiological responses, the system comprising:

[0084] Data acquisition module 110 is used to acquire multi-site electroskin data corresponding to different emotional states;

[0085] Data processing module 120 is used to decompose and extract features from the multi-site electrodermatology data to generate training datasets and test datasets;

[0086] The model building module 130 is used to train a preset prediction model using the training dataset and the test dataset to obtain the final prediction model.

[0087] Prediction module 140 is used to analyze real-time skin conductance data through the final prediction model to obtain the current emotional state;

[0088] The final prediction model is obtained by training a preset prediction model using the training dataset and testing the trained prediction model using the test data.

[0089] Among them, the data acquisition module 110 is attached to the human skin at different locations through a multi-point skin electrosensing device;

[0090] The skin electrosensing device measures the resistance between two electrodes by discharging electrodes and generates a resistance value.

[0091] Based on the resistance values, multi-site electrodermal data corresponding to different emotional states are generated.

[0092] The data processing module 120 decomposes the multi-site electrodermal data into fast-changing electrodermal responses and slow-changing electrodermal responses using a continuous decomposition analysis method.

[0093] Feature extraction of the multi-point electrodermal data includes time-domain feature extraction and frequency-domain feature extraction;

[0094] The data after feature extraction is divided equally to generate training and testing datasets.

[0095] The time-domain features include standard deviation, mean, root mean square, and higher-order statistical features. For short-term event-related response analysis, waveform features of physiological changes are added.

[0096] The frequency domain features include peak frequency and harmonics. Higher-order continuous signal features can be extracted by performing discrete wavelet transformation and empirical mode decomposition on the signal.

[0097] The model building module 130 constructs a classification prediction model for categorized emotion labels, including positive and negative emotions and levels of arousal.

[0098] A regression prediction model is constructed based on continuous emotional ratings.

[0099] The classification and regression prediction models were cross-validated. For data within a single subject, the data were evenly divided into a training set and a test set. The classification and regression models were trained on the training set, and the final prediction model was selected based on the accuracy of the prediction results on the test set.

[0100] Prediction module 140 sends the real-time acquired skin conductance data to the final prediction model;

[0101] The final prediction model is used for analysis and matching to output the predicted emotional state.

[0102] This invention provides an emotion computing system based on multi-site skin physiological responses. By using sensors capable of measuring skin conductance responses at multiple body sites, it induces participants' emotional states with specific emotional material and records the current changes in skin conductance at these sites, establishing an automated emotion recognition method. Compared to traditional single-site skin conductance measurements, the proposed method considers the body distribution of emotion-induced skin conductance responses, which is beneficial for stably and completely discovering the correlation between emotions and skin conductance responses, improving the accuracy of emotion recognition. Furthermore, the wearable method itself has a low workload, contributing to the widespread adoption of everyday wearable emotion measurement, and has significant application value for the development of emotion computing and emotional intelligence.

[0103] Figure 8An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 8 As shown, the electronic device may include a processor 810, a communications interface 820, a memory 830, and a communication bus 840, wherein the processor 810, the communications interface 820, and the memory 830 communicate with each other via the communication bus 840. The processor 810 can call logical instructions in the memory 830 to execute an emotion computing method based on multi-site skin physiological responses, the method including: acquiring multi-site electrodermal data corresponding to different emotional states;

[0104] The multi-site electrodermal data are decomposed and feature extracted to generate training and testing datasets;

[0105] The preset prediction model is trained using the training dataset and the test dataset to obtain the final prediction model;

[0106] The final prediction model is used to analyze real-time skin conductance data to obtain the current emotional state.

[0107] The final prediction model is obtained by training a preset prediction model using the training dataset and testing the trained prediction model using the test data.

[0108] Furthermore, the logical instructions in the aforementioned memory 830 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, essentially, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0109] On the other hand, the present invention also provides a computer program product, the computer program product including a computer program, the computer program being able to be stored on a non-transitory computer-readable storage medium, and when the computer program is executed by a processor, the computer is able to execute an emotion computing method based on multi-site skin physiological response provided by the above methods, the method including: acquiring multi-site skin electrophysiology data corresponding to different emotional states;

[0110] The multi-site electrodermal data are decomposed and feature extracted to generate training and testing datasets;

[0111] The preset prediction model is trained using the training dataset and the test dataset to obtain the final prediction model;

[0112] The final prediction model is used to analyze real-time skin conductance data to obtain the current emotional state.

[0113] The final prediction model is obtained by training a preset prediction model using the training dataset and testing the trained prediction model using the test data.

[0114] In another aspect, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements an emotion computing method based on multi-site skin physiological response provided by the above methods, the method comprising: acquiring multi-site electrodermal data corresponding to different emotional states;

[0115] The multi-site electrodermal data are decomposed and feature extracted to generate training and testing datasets;

[0116] The preset prediction model is trained using the training dataset and the test dataset to obtain the final prediction model;

[0117] The final prediction model is used to analyze real-time skin conductance data to obtain the current emotional state.

[0118] The final prediction model is obtained by training a preset prediction model using the training dataset and testing the trained prediction model using the test data.

[0119] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0120] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0121] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An emotion computing method based on multi-site skin physiological response, characterized in that, The method comprises the following steps: acquiring multi-site skin electric data corresponding to different emotional states; decomposing and extracting features of the multi-site skin electric data to generate a training data set and a test data set; training a preset prediction model through the training data set and the test data set to obtain a final prediction model; analyzing real-time skin electric data through the final prediction model to obtain a current emotional state; wherein the final prediction model is obtained by training the preset prediction model through the training data set and testing the trained prediction model through the test data; the decomposition and feature extraction of the multi-site skin electric data to generate the training data set and the test data set specifically comprises: decomposing the multi-site skin electric data into fast-changing skin electric response and slow-changing skin electric response through continuous decomposition analysis; the feature extraction of the multi-site skin electric data comprises time domain feature extraction and frequency domain feature extraction; the data after decomposition and feature extraction is equally divided to generate the training data set and the test data set. 2.The emotion computing method based on multi-site skin physiological responses according to claim 1, wherein, the acquisition of the multi-site skin electric data corresponding to different emotional states specifically comprises: the multi-site skin electric sensing device is in close contact with the human skin at different parts; the skin electric sensing device measures the resistance between two levels through electrode discharge to generate a resistance value; the multi-site skin electric data corresponding to different emotional states is generated according to the resistance value. 3.The emotion computing method based on multi-site skin physiological responses according to claim 1, wherein, the feature extraction of the multi-site skin electric data comprises time domain feature extraction and frequency domain feature extraction, specifically comprising: the time domain features include standard deviation, mean, root mean square and high-order statistical features, and the waveform features of physiological changes are added for short-term event-related response analysis; the frequency domain features include peak frequency and harmonic, and higher-order continuous signal features can be extracted through discrete wavelet transformation and empirical mode decomposition of the signal. 4.The emotion computing method based on multi-site skin physiological responses according to claim 1, wherein, the training of the preset prediction model through the training data set and the test data set to obtain the final prediction model specifically comprises: constructing a classification prediction model for the categorical emotional labels, including the positive and negative of emotions and the high and low of arousal; constructing a regression prediction model for the continuous score values of emotions; cross-validating the classification prediction model and the regression prediction model, dividing the data of a single subject into a training set and a test set, training the classification and regression models on the training set, and selecting the final prediction model according to the accuracy of the prediction results on the test set. 5.The emotion computing method based on multi-site skin physiological responses according to claim 1, wherein, the analysis of real-time skin electric data through the final prediction model to obtain the current emotional state specifically comprises: sending the real-time acquired skin electric data to the final prediction model; analyzing and matching through the final prediction model to output the predicted emotional state.

6. An emotion computing system based on multi-site skin physiological responses, characterized by, The system comprises: a data acquisition module for acquiring multi-site skin electric data corresponding to different emotional states; a data processing module for decomposing and extracting features of the multi-site skin electric data to generate a training data set and a test data set; a model establishment module for training a preset prediction model through the training data set and the test data set to obtain a final prediction model; The prediction module is configured to analyze real-time electrodermal data by the final prediction model to obtain a current emotional state. The final prediction model is obtained by training a preset prediction model by using the training data set and testing the trained prediction model by using the test data. The data processing module is specifically configured to decompose the multi-site electrodermal data into fast-changing electrodermal responses and slow-changing electrodermal responses by using a continuous decomposition analysis method; extract features of the multi-site electrodermal data, including time-domain feature extraction and frequency-domain feature extraction; and divide the decomposed and feature-extracted data equally to generate the training data set and the test data set.

7. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor executes the program to implement the emotion computing method based on multi-site skin physiological responses according to any one of claims 1 to 5.

8. A non-transitory computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the emotion computing method based on multi-site skin physiological responses according to any one of claims 1 to 5.

9. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to implement the emotion computing method based on multi-site skin physiological responses according to any one of claims 1 to 5.

Images