Emotion evaluation method based on 12-dimensional emotion model

Abstract

The invention relates to an emotion evaluation method based on a 12-dimensional emotion model. The emotion evaluation method comprises the following steps of: acquiring vibration image data of head reflection movement of a tested person by utilizing a vibration image acquisition device; preprocessing the vibration image data, wherein the preprocessing comprises but is not limited to noise removal and feature information extraction; 12 emotion dimension parameters are calculated based on the preprocessed vibration image data; based on the parameters of the 12 emotion dimensions, an emotion evaluation report is generated, and the emotion evaluation report at least comprises the score of each emotion dimension and the description of the overall emotion state. According to the emotion evaluation method based on the 12-dimensional emotion model, the emotion state of the testee can be objectively and accurately recognized and quantified, the limitation that a traditional emotion evaluation method is high in subjectivity and difficult to quantify is overcome, the emotion state can be described more comprehensively and meticulously, and richer information is provided for emotion recognition and intervention.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to an emotion assessment method based on a 12-dimensional emotion model. Background Art

[0002] In recent years, artificial intelligence (AI) technology has flourished around the world and has driven the development of multimodal technology. Multimodal technology refers to the technology of processing and understanding multiple modal information (such as vision, hearing, language, etc.). With breakthroughs in technologies such as deep learning, multimodal learning has become a new research trend and has made significant progress in many fields, such as emotion recognition. Multimodal technology can identify emotional states more comprehensively and accurately by comprehensively analyzing multiple modal information, such as facial expressions, voice intonation, body movements, etc., showing great application potential.

[0003] Multimodal technology has also brought new development opportunities for psychological assessment technology. Psychological assessment technology refers to the technology of evaluating individual psychological characteristics and behaviors through scientific methods, such as intelligence tests, personality tests, emotional tests, etc., which are widely used in mental health, education, human resource management and other fields. Traditional psychological assessment technology mainly relies on highly subjective methods such as questionnaires and interviews, which have limitations such as being easily affected by subjective factors and difficult to quantify. The introduction of artificial intelligence and multimodal technology provides new ideas for improving the objectivity and accuracy of psychological assessment.

[0004] As an important direction of psychological assessment, emotion assessment also faces the challenge of technological innovation. Traditional emotion assessment methods mainly rely on single-modal information such as facial expression recognition and speech analysis, which are limited in accuracy and reliability. In order to overcome these limitations, researchers have begun to explore the use of multimodal data for emotion assessment, such as combining multiple information such as facial expressions, voice intonation, and physiological signals to identify emotional states.

[0005] However, existing multimodal emotion assessment techniques still have the following limitations:

[0006] (I) Limitations of data fusion methods: Most existing multimodal data fusion methods remain at the simple splicing level of features, and it is difficult to fully utilize the complementarity and correlation between different modal information.

[0007] (II) Limitations of emotion models: Most existing emotion models are based on discrete emotion categories, such as joy, anger, sadness, and happiness, which make it difficult to describe the complexity and continuity of emotions.

[0008] (III) Limitations of application scenarios: Existing emotion assessment technologies are mostly used in laboratory environments and are difficult to meet the complex and changing needs in real scenarios. Summary of the invention

[0009] In order to solve the above problems, the present invention provides an emotion assessment method based on a 12-dimensional emotion model that can identify and quantify emotional states more objectively, accurately and comprehensively.

[0010] In order to achieve the above object, the emotion assessment method based on the 12-dimensional emotion model designed by the present invention comprises the following steps:

[0011] S1. Collecting vibration image data of the test subject's head reflex motion using a vibration image acquisition device;

[0012] S2. Preprocessing the vibration image data, including but not limited to removing noise and extracting feature information;

[0013] S3. Based on the pre-processed vibration image data, 12 emotion dimension parameters are calculated, and the 12 emotion dimensions include: gentle and friendly, relaxed and peaceful, leisure and comfortable, firm and trusting, balanced and stable, charming and confident, energetic and energetic, self-regulation level, encouraging and progressive, nervous and sensitive, happy and cheerful, and happy and joyful;

[0014] S4. Based on the parameters of the 12 emotion dimensions, generate an emotion assessment report, wherein the emotion assessment report at least includes a score of each emotion dimension and a description of the overall emotion state.

[0015] Optionally, the vibration image acquisition device includes multiple cameras, a signal processor, a data storage device and a display device, and the multiple cameras are distributed around the head of the test subject to form a closed head reflex motion capture network; the signal processor is electrically connected to the multiple cameras for processing the vibration image data captured by the cameras; the data storage device is electrically connected to the signal processor for storing the vibration image data processed by the signal processor.

[0016] Optionally, the vibration image acquisition device further comprises a display device, which is electrically connected to the signal processor and is used for displaying the tester's head reflection motion signal in real time.

[0017] Optionally, in step S1, the vibration image data is collected in a shadowless light source environment.

[0018] Optionally, in step S4, the display form of the emotion assessment report includes a chart form or a text form.

[0019] Optionally, in step S3:

[0020] The gentleness and friendliness parameter calculation process includes:

[0021] S310. Perform frequency domain analysis on the preprocessed vibration image data to generate a vibration image frequency histogram to obtain vibration data within a frequency range below 5 Hz;

[0022] S311. Based on the vibration data in the frequency range below 5 Hz, calculate the maximum value G in the vibration image frequency histogram m and standard deviation τ to calculate the gentleness and friendliness parameter E2. The calculation formula of the gentleness and friendliness parameter E2 is as follows:

[0023]

[0024] Among them, G in It is the average value of all vibration data within the frequency range below 5 Hz;

[0025] The relaxation and parameter calculation process includes:

[0026] S320. Perform frequency domain analysis on the pre-processed vibration image data to generate a vibration spectrum and obtain the maximum value of the frequency g max ;

[0027] S321. Based on the vibration spectrum, calculate the probability density function Q related to the frequency i (g), the probability density function Qi(g) represents the spectral power density obtained by Fourier transforming the vibration image data at the vibration image frequency g;

[0028] S322. Frequency-dependent probability density function Q i (g) and the maximum value of the frequency g max , calculate the relaxation peace parameter E4, the calculation formula of the relaxation peace parameter E4 is as follows:

[0029]

[0030] The leisure comfort parameter calculation process includes:

[0031] S330. Based on the vibration image data, obtain a horizontal motion data set X and a vertical motion data set Y of the head of the test subject, each set containing m elements;

[0032] S331. For each set, obtain the upper value and lower value of each element therein, where the upper value and lower value of the jth element in set X are X respectively. U and X D , the upper and lower values ​​of the jth element in the set Y are Y U and Y D , j is the index variable, from 1 to m;

[0033] S332. Get the maximum value X of all elements in set X max The maximum value Y of all elements in the set Y max ;

[0034] S333.Based on X U , X D , Y U , Y D , X max , Y max and m, calculate the leisure comfort parameter E5, the calculation formula of the leisure comfort parameter E5 is as follows:

[0035]

[0036] The process of calculating the trust parameter includes:

[0037] S340. Obtaining the aggressiveness level parameter C1 which is negatively correlated with the gentleness and friendliness parameter E2;

[0038] S341. Obtaining a stress level parameter C2 that is negatively correlated with the leisure comfort parameter E5;

[0039] S342. Obtaining the tension and anxiety level parameter C3 which is negatively correlated with the relaxation and peace parameter E4;

[0040] S343. Based on the aggressiveness level parameter C1, the stress level parameter C2 and the tension and anxiety level parameter C3, the confidence parameter R4 is calculated. The calculation formula of the confidence parameter R4 is as follows:

[0041]

[0042] The equilibrium stability parameter calculation process includes:

[0043] S350. Calculate the average values ​​of the 11 emotional parameters of gentleness and friendliness, relaxation and peace, leisure and comfort, firmness and trust, charm and confidence, vitality and energy, self-regulation level, encouragement and progress, neurotic sensitivity, happiness and joy, and happiness and joy within the predetermined time;

[0044] S351. Calculate the standard deviation of 11 emotion parameters in the same time period respectively;

[0045] S352. The ratio of the standard deviation of each emotion parameter to its mean is taken as the coefficient of variation Cl of the emotion parameter;

[0046] S353. Calculate the sum of the variation values ​​Wb of each emotion parameter based on the coefficient of variation Cl, and calculate the balance stability parameter F5. The calculation formula of the balance stability parameter F5 is as follows:

[0047] F5 = (100-2Wb)%;

[0048] The charm confidence parameter calculation process includes:

[0049] S360. Based on the vibration image data, obtain the difference between the average vibration amplitudes of the left and right sides of the vibration image of each facial contour line | X li -X ri ∣;

[0050] S361. Based on the vibration image data, obtain the maximum frequency difference between the vibration amplitudes on the left and right sides of the vibration image of each facial contour line | Y li -Y ri ∣;

[0051] S362. Based on |X li -X ri ∣ and ∣Y li -Y ri ∣, calculate the charm confidence parameter F6, the calculation formula of the charm confidence parameter F6 is as follows:

[0052]

[0053] Where M is the number of lines of the vibration image, j is an index variable ranging from 1 to M; li and ri represent the vibration amplitude of the i-th line on the left and the i-th line on the right of the vibration image, respectively;

[0054] The vitality energy parameter calculation process includes:

[0055] S370. Obtaining image vibration frequency histogram;

[0056] S371. Extract the maximum density value P and the standard deviation τ in the image vibration frequency histogram;

[0057] S372. Based on the maximum density value P and the standard deviation τ, the vitality energy R7 is calculated. The calculation formula of the vitality energy R7 is as follows:

[0058]

[0059] Among them, Q sp is the maximum value of the vibration image input frequency;

[0060] The self-regulation level parameter calculation process includes:

[0061] S380. Obtain the difference dF5 of the equilibrium stability parameter F5;

[0062] S381. Obtain the difference dF6 of the charm confidence parameter F6;

[0063] S382. Based on dF5, dF6, F5 and F6, calculate the self-regulation level parameter R8, the calculation formula of the self-regulation level parameter R8 is as follows:

[0064]

[0065] The encouraging progress parameter calculation process includes:

[0066] S390. Obtain the time series data G1 of the gentleness and friendliness parameter E2;

[0067] S391. Perform periodic analysis on the time series data G1 to determine whether the gentleness and friendliness parameter E2 has periodic changes;

[0068] S392. If the gentleness and friendliness parameter E2 has periodic changes, obtain the maximum value related to the time series data G1 or the time M under specific conditions x (G1), and the total time B;

[0069] S393. Based on time M x (G1) and time B, calculate the encouraging progress parameter R9, the calculation formula of the encouraging progress parameter R9 is as follows:

[0070]

[0071] The neuroticism sensitivity parameter calculation process includes:

[0072] S3100. Get the encouraging progress parameter R9;

[0073] S3101. Based on the encouragement and progress parameter R9, the neurotic sensitivity parameter R10 is calculated. The calculation formula of the neurotic sensitivity parameter R10 is as follows:

[0074] R10 = 10τ(R9);

[0075] The happiness parameter calculation process includes:

[0076] S3110. Obtain the peak value P corresponding to the maximum frequency in the vibration image frequency histogram;

[0077] S3111. Based on the standard deviation τ and the peak value P, the happiness parameter R11 is calculated. The calculation formula of the happiness parameter R11 is as follows:

[0078]

[0079] The happiness parameter calculation process includes:

[0080] S3120. Calculate information efficiency X and energy consumption Y based on vibration image data;

[0081] S3121. Calculate the change in information efficiency dU; calculate the change in energy consumption dV;

[0082] S3122. Based on the information efficiency X, energy consumption Y, the change in information efficiency dU and the change in energy consumption dV, calculate the happiness parameter R12. The calculation formula of the happiness parameter R12 is as follows:

[0083]

[0084] The emotion assessment method designed by the present invention based on the 12-dimensional emotion model can objectively and accurately identify and quantify the emotional state of the test subject, overcoming the limitations of traditional emotion assessment methods that are highly subjective and difficult to quantify, that is, it can more comprehensively and meticulously portray the emotional state, providing richer information for emotion recognition and intervention. In addition, the method of the present invention is simple to operate and easy to implement, and can be applied in a variety of scenarios, such as mental health assessment, education and training, human-computer interaction, etc., and has broad application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0085] Figure 1 It is a flowchart of the emotion assessment method based on the 12-dimensional emotion model provided in an embodiment of the present application. DETAILED DESCRIPTION

[0086] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0087] Example 1

[0088] like Figure 1 As shown, the emotion assessment method based on the 12-dimensional emotion model described in this embodiment includes the following steps:

[0089] S1. Use a vibration image acquisition device to collect vibration image data of the test subject's head reflex movement.

[0090] Preferably, the vibration image acquisition device includes multiple cameras, a signal processor, a data storage device and a display device, and the multiple cameras are distributed around the head of the test subject to form a closed head reflection motion capture network; the signal processor is electrically connected to the multiple cameras for processing the vibration image data captured by the cameras; the data storage device is electrically connected to the signal processor for storing the vibration image data processed by the signal processor.

[0091] Specifically, the test subject wears a vibration image acquisition device, which includes multiple cameras, for example, 8 or more, evenly distributed around the test subject's head to form a closed head reflection motion capture network. Each camera is electrically connected to a signal processor to collect real-time vibration image data of the reflection motion of the test subject's head under a shadowless light source.

[0092] Preferably, the vibration image acquisition device further comprises a display device, which is electrically connected to the signal processor and is used to display the tester's head reflection motion signal in real time, such as a liquid crystal display screen, which can display the head motion signal in real time, making it convenient for the operator to observe and monitor the data acquisition process.

[0093] S2. Preprocessing the vibration image data, wherein the preprocessing includes but is not limited to removing noise and extracting feature information.

[0094] Specifically, vibration image data usually contains interference signals such as environmental noise and equipment noise, which need to be removed using a noise reduction algorithm. For example, a low-pass filter can be used to remove high-frequency noise, and a band-pass filter can be used to retain signals within a specific frequency range; then, feature information related to emotions, such as frequency, amplitude, phase, energy, etc., is extracted from the pre-processed vibration image data. In this embodiment, existing time-frequency analysis, statistical analysis, and other methods can be used to extract features.

[0095] After preprocessing, the noise of the vibration image data was suppressed, the signal strength was enhanced, and key feature information was extracted, laying the foundation for subsequent emotion assessment analysis; finally, the preprocessed data was transmitted to the data storage device for storage.

[0096] S3. Based on the pre-processed vibration image data, 12 emotion dimension parameters are calculated, and the 12 emotion dimensions include: gentleness and friendliness, relaxation and peace, leisure and comfort, firmness and trust, balance and stability, charm and confidence, vitality and energy, self-regulation level, encouragement and progress, neurotic sensitivity, happiness and joy, and happiness and joy. The detailed calculation method of each parameter will be introduced in detail in the subsequent section.

[0097] S4. Generate an emotion assessment report based on the parameters of the 12 emotion dimensions, wherein the emotion assessment report at least includes a score of each emotion dimension and a description of the overall emotion state. In this embodiment, the emotion assessment report is presented in a graphical form or in a textual form.

[0098] Specifically, taking a radar chart as an example, the chart shows the scores of 12 emotional dimensions. Each dimension occupies a direction. The higher the score, the farther the corresponding point is from the center. That is, through the radar chart, you can intuitively see which emotional dimensions the tester scored higher and which emotional dimensions scored lower. For example, the tester showed high vitality and encouragement during the test, but the level of tension and anxiety was also relatively high, which may be a manifestation of excitement and tension in the test situation.

[0099] The obtained emotional assessment report can help psychologists understand the patient's emotional state more comprehensively, assist in diagnosing emotional disorders such as depression and anxiety, and develop more effective treatment plans. Alternatively, through emotional assessment, we can understand consumers' true emotional reactions to products or services, such as like, disgust, excitement, etc., so as to improve product design and optimize marketing strategies.

[0100] Below, the calculation of the 12 emotion dimension parameters provided in the embodiment of the present application will be described in detail.

[0101] The first emotional dimension: Gentleness / Friendliness (peaceful) is the opposite indicator of aggression (aggression).

[0102] The calculation process of the gentleness parameter E2 is as follows:

[0103] S310. Perform frequency domain analysis on the preprocessed vibration image data to generate a vibration image frequency histogram to obtain vibration data within a frequency range below 5 Hz; the histogram uses frequency as the horizontal coordinate and vibration amplitude or power as the vertical coordinate to intuitively display the intensity distribution of vibration signals of different frequencies during head movement.

[0104] S311. Based on the vibration data in the frequency range below 5 Hz, calculate the maximum value G in the vibration image frequency histogram m and standard deviation τ to calculate the gentleness and friendliness parameter E2. The calculation formula of the gentleness and friendliness parameter E2 is as follows:

[0105]

[0106] Among them, G in is the average value of all vibration data within the frequency range below 5 Hz; and the standard deviation τ represents the degree of deviation between the values ​​in the data set and the average value Gin.

[0107] In this way, the gentleness and friendliness parameter is quantified into a specific value, which can be compared and analyzed, making it easier to conduct data statistics and research. mreflects the peak value of the vibration intensity and can capture the intensity of the emotion; the standard deviation τ reflects the discreteness of the vibration data and can capture the fluctuation of the emotion; the average value G in It reflects the overall level of vibration data and can be used as a reference benchmark. For example, if the E2 value is high, it means that the subject shows higher intensity and fluctuation in low-frequency vibration, suggesting a higher level of gentleness and friendliness.

[0108] The second emotional dimension: relaxation is the reverse derivation of tension. The relaxation parameter calculation process includes:

[0109] S320. Perform frequency domain analysis on the pre-processed vibration image data to generate a vibration spectrum and obtain the maximum value of the frequency g max ; The spectrum uses frequency as the horizontal axis and the power density of the vibration signal as the vertical axis, reflecting the intensity distribution of vibration signals of different frequencies during head movement.

[0110] S321. Based on the vibration spectrum, calculate the probability density function Q related to the frequency i (g), the probability density function Qi(g) represents the spectral power density obtained by Fourier transforming the vibration image data at the vibration image frequency g; that is, Q i The larger the value of (g), the stronger the vibration signal at frequency g.

[0111] S322. Frequency-dependent probability density function Q i (g) and the maximum value of the frequency g max , calculate the relaxation peace parameter E4, the calculation formula of the relaxation peace parameter E4 is as follows:

[0112]

[0113] In this way, the relaxation parameter is quantified into a specific value, which can be compared and analyzed, making it easier to conduct data statistics and research. i (g) reflects the intensity distribution of vibration signals of different frequencies, which can capture the impact of different frequency components on relaxation and peaceful emotions; and the maximum frequency g max It determines the frequency range considered when calculating E4 and can be adjusted according to actual conditions.

[0114] For example, considering that frequencies above 30 Hz are associated with tension and anxiety, in this embodiment, g maxSet to 30Hz, at this time, the E4 value reflects the proportion of low-frequency vibration energy below 30Hz. The higher the E4 value, the higher the proportion of low-frequency vibration energy below 30Hz, suggesting a higher level of relaxation and peace.

[0115] The third emotional dimension: relaxation is the reverse derivation of stress. The calculation process of the relaxation parameter includes:

[0116] S330. Based on the vibration image data, obtain the horizontal motion data set X and the vertical motion data set Y of the subject's head respectively, each set contains m elements; each set contains m data points, representing the data collected within a period of time.

[0117] S331. For each set, obtain the upper value and lower value of each element therein, where the upper value and lower value of the jth element in set X are X respectively. U and X D , the upper and lower values ​​of the jth element in the set Y are Y U and Y D , j is the index variable, from 1 to m.

[0118] S332. Get the maximum value X of all elements in set X max The maximum value Y of all elements in the set Y max .

[0119] S333.Based on X U , X D , Y U , Y D , X max , Y max and m, calculate the leisure comfort parameter E5, the calculation formula of the leisure comfort parameter E5 is as follows:

[0120]

[0121] In this way, by calculating the asymmetry of the horizontal and vertical movement ranges of the test subject's head, we can obtain the leisure comfort parameter. The smaller the E5 value, the more symmetrical the head movement is, and the more relaxed and comfortable the test subject is; the larger the E5 value, the more asymmetric the head movement is, and the more stressed the test subject is.

[0122] The fourth emotional dimension: confidence / trust is the reverse derivation of suspicion / doubt. The confidence / trust parameter calculation process includes:

[0123] S340. Obtain the aggressiveness level parameter C1 that is negatively correlated with the gentleness and friendliness parameter E2; since aggressiveness and gentleness and friendliness are inverse indicators, the formula C1 = 100-E2 can be used for calculation.

[0124] S341. Obtain the stress level parameter C2 which is negatively correlated with the leisure comfort parameter E5; since stress and leisure comfort are inverse indicators, the formula C2=100-E5 can be used for calculation.

[0125] S342. Obtain the tension and anxiety level parameter C3 which is negatively correlated with the relaxation and peace parameter E4; since tension and anxiety and relaxation and peace are inverse indicators, the formula C3=100-E4 can be used for calculation.

[0126] S343. Based on the aggressiveness level parameter C1, the stress level parameter C2 and the tension and anxiety level parameter C3, the confidence parameter R4 is calculated. The calculation formula of the confidence parameter R4 is as follows:

[0127]

[0128] In this way, by incorporating the three negative emotion indicators of aggression, stress, and tension and anxiety into the calculation, the user's negative emotion level can be more comprehensively reflected, thereby more accurately inferring their level of firm trust.

[0129] The fifth emotional dimension: balance is a manifestation of emotional stability. The balance parameter calculation process includes:

[0130] S350. Calculate the average values ​​of the 11 emotional parameters of gentleness and friendliness, relaxation and peace, leisure and comfort, firmness and trust, charm and confidence, vitality and energy, self-regulation level, encouragement and progress, neurotic sensitivity, happiness and joy, and happiness and joy within the predetermined time;

[0131] S351. Calculate the standard deviation of 11 emotion parameters in the same time period respectively;

[0132] S352. The ratio of the standard deviation of each emotion parameter to its mean is taken as the coefficient of variation Cl of the emotion parameter;

[0133] S353. Calculate the sum of the variation values ​​Wb of each emotion parameter based on the coefficient of variation Cl, and calculate the balance stability parameter F5. The calculation formula of the balance stability parameter F5 is as follows:

[0134] F5=(100-2Wb)%.

[0135] In this way, the coefficient of variation Cl of each emotional parameter can more stably reflect the characteristics of the parameter. A smaller coefficient of variation Cl represents the stability of the emotional parameter per unit time. By calculating the coefficient of variation of individual emotional parameters, the user's personalized emotional fluctuation characteristics can be captured, avoiding the bias of subjective evaluation and making the evaluation results more objective and reliable.

[0136] For example, assuming that the coefficient of variation Cl of a user's "happy and cheerful" emotional parameter in the past week is small, this indicates that the user's emotional experience in this week is relatively stable, and the degree of happiness and cheerfulness fluctuates less; conversely, if another user's "neurotic and sensitive" emotional parameter has a larger coefficient of variation Cl in the same time period, it indicates that the user's "neurotic and sensitive" emotions fluctuate greatly and his emotional state is not very stable.

[0137] It should be noted that since the calculation of the balance stability parameter depends on the calculation results of the other 11 emotion parameters, in actual implementation, the other 11 emotion parameters can be calculated first, and then the balance stability parameter F5 can be calculated, which can avoid repeated calculations and improve efficiency.

[0138] The sixth emotional dimension: Charm is the cognitive feedback of one’s current state. The charm and confidence parameter calculation process includes:

[0139] S360. Based on the vibration image data, obtain the difference between the average vibration amplitudes of the left and right sides of the vibration image of each facial contour line | X li -X ri ∣;

[0140] S361. Based on the vibration image data, obtain the maximum frequency difference between the vibration amplitudes on the left and right sides of the vibration image of each facial contour line | Y li -Y ri ∣;

[0141] S362. Based on |X li -X ri ∣ and ∣Y li -Y ri ∣, calculate the charm confidence parameter F6, the calculation formula of the charm confidence parameter F6 is as follows:

[0142]

[0143] Among them, M is the number of lines of the vibration image, j is an index variable ranging from 1 to M; li and ri represent the vibration amplitude of the i-th line on the left and the i-th line on the right of the vibration image respectively.

[0144] In this way, by analyzing the frequency and amplitude of the subject's head and facial muscle micro-movements, the charm confidence parameter F6 can be calculated. The higher the frequency and amplitude of the movement, the better the symmetry, and the higher the F6 value, indicating that the test subject has a higher level of charm confidence.

[0145] For example, if the test subject's facial expression is relaxed and natural, and the frequency and amplitude of muscle movements on the left and right sides are highly consistent, then |X li -X ri | and |Y li- Y ri | will be very small, and the final calculated F6 value will be high, indicating that the test subject has a high level of charm and confidence. On the contrary, if the test subject's facial expression is tense and stiff, and the muscle movements on the left and right sides are asymmetrical, the F6 value will be low, indicating a low level of charm and confidence.

[0146] The seventh emotional dimension: Vitality\Energy is a reflection of the current psychological energy. The calculation process of the Vitality\Energy parameter includes:

[0147] S370. Obtaining image vibration frequency histogram;

[0148] S371. Extract the maximum density value P and the standard deviation τ in the image vibration frequency histogram;

[0149] S372. Based on the maximum density value P and the standard deviation τ, the vitality energy R7 is calculated. The calculation formula of the vitality energy R7 is as follows:

[0150]

[0151] Among them, Q sp is the maximum value of the vibration image input frequency.

[0152] In this way, the vital energy parameter characterizes the difference between the maximum density value of the vibration frequency of the human head and face and the frequency standard deviation.

[0153] Specifically, the greater the maximum value of the vibration frequency density, the higher the intensity of the main vibration frequency of the head and face. At the same time, if the standard deviation or dispersion of the vibration frequency is lower, it means that the distribution of the vibration frequency is more concentrated and the person's state is more stable. Combining these two aspects of information, suppose there are two test subjects A and B, whose maximum density values ​​of vibration frequency are relatively high, but the standard deviation of A's vibration frequency is significantly lower than that of B, which means that A's vibration frequency is more concentrated and the state is more stable, so A's vitality level is higher than B.

[0154] The eighth emotional dimension: self-regulation is the subject's ability to restore positive emotions. The self-regulation parameter calculation process includes:

[0155] S380. Obtain the difference dF5 of the equilibrium stability parameter F5;

[0156] S381. Obtain the difference dF6 of the charm confidence parameter F6;

[0157] S382. Based on dF5, dF6, F5 and F6, calculate the self-regulation level parameter R8, the calculation formula of the self-regulation level parameter R8 is as follows:

[0158]

[0159] Specifically, the calculation of the self-regulation level parameter R8 does not directly use the values ​​of the balance stability parameter F5 and the charm confidence parameter F6, but uses their changes in adjacent time periods, that is, the difference dF5 and dF6. The advantage of this is that it does not only rely on the emotional state at a single moment, but takes into account the changing trend of emotional parameters over a period of time, which can better reflect the dynamic change process of the self-regulation ability of the test subject.

[0160] For example, suppose that when a subject faces a stressful situation, his balance stability parameter F5 and charm confidence parameter F6 both decrease, but in the following period of time, the difference between these two parameters dF5 and dF6 both show a positive growth trend, which indicates that the subject can gradually overcome the influence of negative emotions and restore a positive emotional state, and his self-regulation ability is strong.

[0161] The ninth emotional dimension: confidence is the reverse derivation of inhibition. The confidence parameter calculation process includes:

[0162] S390. Obtain the time series data G1 of the gentleness and friendliness parameter E2;

[0163] S391. Perform periodic analysis on the time series data G1 to determine whether the gentleness and friendliness parameter E2 has periodic changes;

[0164] S392. If the gentleness and friendliness parameter E2 has periodic changes, obtain the maximum value related to the time series data G1 or the time M under specific conditions x (G1), and the total time B;

[0165] S393. Based on time M x (G1) and time B, calculate the encouraging progress parameter R9, the calculation formula of the encouraging progress parameter R9 is as follows:

[0166]

[0167] In this embodiment, encouragement / progress and inhibition / repression are opposite emotional dimensions, and the inhibition / repression parameter E9 is defined as the average period of the gentleness and friendliness parameter E2, which is used to measure the degree of inhibition and repression of an individual, wherein the average period refers to the average length of time that the gentleness and friendliness parameter E2 presents periodic changes.

[0168] Preferably, the default value of the suppression parameter E9 is 100 frames, corresponding to a reaction time of 0.1 seconds, which means that if the value of the suppression parameter E9 fluctuates within a 10% range, it indicates that the displacement of the center of gravity of the human head undergoes periodic vibrations approximately 10 times per second.

[0169] It should be noted that the actual value of the inhibition\suppression parameter E9 should be determined by the maximum frame rate of the vibration image acquisition device, and according to the Shannon sampling theorem (Nyquist-Shannon-Kotelnikov theorem), the frame rate for calculating this parameter should be at least 25 frames per second (fps).

[0170] Therefore, by analyzing the average period of the gentleness and friendliness parameter E2, we can infer the degree of inhibition and repression of the individual and then evaluate his or her level of encouragement and progress.

[0171] The tenth emotional dimension: neuroticism is the degree of the subject's reaction to people or things around him. The neuroticism parameter calculation process includes:

[0172] S3100. Get the encouraging progress parameter R9;

[0173] S3101. Based on the encouragement and progress parameter R9, the neurotic sensitivity parameter R10 is calculated. The calculation formula of the neurotic sensitivity parameter R10 is as follows:

[0174] R10=10τ(R9).

[0175] Specifically, neurotic and sensitive people usually have greater emotional fluctuations, and the encouragement and progress parameter R9 reflects the degree of fluctuation of this parameter over a period of time compared to the standard deviation τ. For example, assuming the changes in the encouragement and progress parameter R9 of two diagonal users within a week, the R9 value of user A fluctuates greatly and the standard deviation τ is higher; while the R9 value of user B is relatively stable and the standard deviation τ is lower. It can be inferred that user A is more susceptible to the external environment than user B and has a higher degree of neurotic sensitivity. Therefore, using the parameter R9 and the standard deviation τ to calculate the neurotic sensitivity parameter R10 not only simplifies the calculation process and improves efficiency, but also can more objectively reflect the user's neurotic sensitivity.

[0176] The eleventh emotional dimension: cheerfulness is the reverse derivation of depression. The cheerfulness parameter calculation process includes:

[0177] S3110. Obtain the peak value P corresponding to the maximum frequency in the vibration image frequency histogram;

[0178] S3111. Based on the standard deviation τ and the peak value P, the happiness parameter R11 is calculated. The calculation formula of the happiness parameter R11 is as follows:

[0179]

[0180] Specifically, the calculation of the happiness parameter takes into account the threshold of the standard deviation of the vibration frequency, which is set at 0.5Hz. This is because, generally speaking, a vibration frequency standard deviation exceeding 0.5Hz indicates abnormal fluctuations in psychophysiology, which may indicate various types of psychophysiological disorders. For example, people with anxiety or depression often have a vibration frequency standard deviation higher than 0.5Hz, reflecting their large mood swings and abnormal nervous system activity. Therefore, using 0.5Hz as a threshold can help us more accurately assess an individual's happiness and take potential psychophysiological problems into consideration.

[0181] The twelfth emotional dimension: happiness is the subject's current level of pleasure. The happiness parameter calculation process includes:

[0182] S3120. Calculate information efficiency X and energy consumption Y based on vibration image data;

[0183] S3121. Calculate the change in information efficiency dU; calculate the change in energy consumption dV;

[0184] S3122. Based on the information efficiency X, energy consumption Y, the change in information efficiency dU and the change in energy consumption dV, calculate the happiness parameter R12. The calculation formula of the happiness parameter R12 is as follows:

[0185]

[0186] In this way, not only the current information efficiency X and energy consumption Y are taken into account, but also their changes dU and dV, which can more comprehensively reflect the user's working status and psychological feelings. For example, even if the current information efficiency is high, if the energy consumption continues to increase, or the information efficiency shows a downward trend, the user's happiness may also decrease.

[0187] The emotion assessment method based on the 12-dimensional emotion model provided in this embodiment can objectively and accurately identify and quantify the emotional state of the test subject, overcoming the limitations of traditional emotion assessment methods that are highly subjective and difficult to quantify, that is, it can more comprehensively and meticulously portray the emotional state, providing richer information for emotion recognition and intervention. In addition, the method of the present invention is simple to operate and easy to implement, and can be applied in a variety of scenarios, such as mental health assessment, education and training, human-computer interaction, etc., and has broad application prospects.

[0188] In the description of the present invention, it should be noted that the terms "vertical", "up", "down", "horizontal", etc. indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention.

[0189] In the description of the present invention, it is also necessary to explain that, unless otherwise clearly specified and limited, the terms "set", "install", "connect", and "connect" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two elements. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0190] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein by equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for emotion assessment based on a 12-dimensional emotion model, characterized in that: The following steps are involved: S1. Using a vibration image acquisition device to collect vibration image data of the test subject's head reflex movement; S2. Preprocessing the vibration image data, including but not limited to removing noise and extracting feature information; S3. Based on the pre-processed vibration image data, 12 emotion dimension parameters are calculated, and the 12 emotion dimensions include: gentle and friendly, relaxed and peaceful, leisure and comfortable, firm and trusting, balanced and stable, charming and confident, energetic and energetic, self-regulation level, encouraging and progressive, nervous and sensitive, happy and cheerful, and happy and joyful; S4. Based on the parameters of the 12 emotion dimensions, generate an emotion assessment report, wherein the emotion assessment report at least includes a score of each emotion dimension and a description of the overall emotion state.

2. The emotion assessment method based on the 12-dimensional emotion model according to claim 1, characterized in that: The vibration image acquisition device includes multiple cameras, a signal processor, a data storage device and a display device. The multiple cameras are distributed around the head of the test subject to form a closed head reflection motion capture network; the signal processor is electrically connected to the multiple cameras and is used to process the vibration image data captured by the cameras; the data storage device is electrically connected to the signal processor and is used to store the vibration image data processed by the signal processor.

3. The emotion assessment method based on the 12-dimensional emotion model according to claim 2, characterized in that: The vibration image acquisition device also includes a display device, which is electrically connected to the signal processor and is used to display the tester's head reflection motion signal in real time.

4. The emotion assessment method based on the 12-dimensional emotion model according to claim 1, characterized in that: In step S1 , the vibration image data is collected in a shadowless light source environment.

5. The emotion assessment method based on the 12-dimensional emotion model according to claim 1, characterized in that: In step S4, the emotion assessment report is presented in a graphical or textual form.

6. The emotion assessment method based on the 12-dimensional emotion model according to any one of claims 1 to 5, characterized in that: In step S3: The gentleness and friendliness parameter calculation process includes: S310. Perform frequency domain analysis on the preprocessed vibration image data to generate a vibration image frequency histogram to obtain vibration data within a frequency range below 5 Hz; S311. Based on the vibration data in the frequency range below 5 Hz, calculate the maximum value G in the vibration image frequency histogram m and standard deviation τ to calculate the gentleness and friendliness parameter E2. The calculation formula of the gentleness and friendliness parameter E2 is as follows: Among them, G in It is the average value of all vibration data within the frequency range below 5 Hz; The relaxation and parameter calculation process includes: S320. Perform frequency domain analysis on the pre-processed vibration image data to generate a vibration spectrum and obtain the maximum value of the frequency g max ; S321. Based on the vibration spectrum, calculate the probability density function Q related to the frequency i (g), the probability density function Qi(g) represents the spectral power density obtained by Fourier transforming the vibration image data at the vibration image frequency g; S322. Frequency-dependent probability density function Q i (g) and the maximum value of the frequency g max , calculate the relaxation peace parameter E4, the calculation formula of the relaxation peace parameter E4 is as follows: The leisure comfort parameter calculation process includes: S330. Based on the vibration image data, obtain a horizontal motion data set X and a vertical motion data set Y of the head of the test subject, each set containing m elements; S331. For each set, obtain the upper value and lower value of each element therein, where the upper value and lower value of the jth element in set X are X respectively. U and X D , the upper and lower values ​​of the jth element in the set Y are Y U and Y D , j is the index variable, from 1 to m; S332. Get the maximum value X of all elements in set X max The maximum value Y of all elements in the set Y max ; S333.Based on X U , X D , Y U , Y D , X max , Y max and m, calculate the leisure comfort parameter E5, the calculation formula of the leisure comfort parameter E5 is as follows: The process of calculating the trust parameter includes: S340. Obtaining the aggressiveness level parameter C1 which is negatively correlated with the gentleness and friendliness parameter E2; S341. Obtaining a stress level parameter C2 that is negatively correlated with the leisure comfort parameter E5; S342. Obtaining the tension and anxiety level parameter C3 which is negatively correlated with the relaxation and peace parameter E4; S343. Based on the aggressiveness level parameter C1, the stress level parameter C2 and the tension and anxiety level parameter C3, the confidence parameter R4 is calculated. The calculation formula of the confidence parameter R4 is as follows: The equilibrium stability parameter calculation process includes: S350. Calculate the average values ​​of the 11 emotional parameters of gentleness and friendliness, relaxation and peace, leisure and comfort, firmness and trust, charm and confidence, vitality and energy, self-regulation level, encouragement and progress, neurotic sensitivity, happiness and joy, and happiness and joy within the predetermined time; S351. Calculate the standard deviation of 11 emotion parameters in the same time period respectively; S352. The ratio of the standard deviation of each emotion parameter to its mean is taken as the coefficient of variation Cl of the emotion parameter; S353. Calculate the sum of the variation values ​​Wb of each emotion parameter based on the coefficient of variation Cl, and calculate the balance stability parameter F5. The calculation formula of the balance stability parameter F5 is as follows: F5 = (100-2Wb)%; The charm confidence parameter calculation process includes: S360. Based on the vibration image data, obtain the difference between the average vibration amplitudes of the left and right sides of the vibration image of each facial contour line | X li -X ri ∣; S361. Based on the vibration image data, obtain the maximum frequency difference between the vibration amplitudes on the left and right sides of the vibration image of each facial contour line | Y li -Y ri ∣; S362. Based on |X li -X ri ∣ and ∣Y li -Y ri ∣, calculate the charm confidence parameter F6, the calculation formula of the charm confidence parameter F6 is as follows: Where M is the number of lines of the vibration image, j is an index variable ranging from 1 to M; li and ri represent the vibration amplitude of the i-th line on the left and the i-th line on the right of the vibration image, respectively; The vitality energy parameter calculation process includes: S370. Obtaining image vibration frequency histogram; S371. Extract the maximum density value P and the standard deviation τ in the image vibration frequency histogram; S372. Based on the maximum density value P and the standard deviation τ, the vitality energy R7 is calculated. The calculation formula of the vitality energy R7 is as follows: Among them, Q sp is the maximum value of the vibration image input frequency; The self-regulation level parameter calculation process includes: S380. Obtain the difference dF5 of the equilibrium stability parameter F5; S381. Obtain the difference dF6 of the charm confidence parameter F6; S382. Based on dF5, dF6, F5 and F6, calculate the self-regulation level parameter R8, the calculation formula of the self-regulation level parameter R8 is as follows: The encouraging progress parameter calculation process includes: S390. Obtain the time series data G1 of the gentleness and friendliness parameter E2; S391. Perform periodic analysis on the time series data G1 to determine whether the gentleness and friendliness parameter E2 has periodic changes; S392. If the gentleness and friendliness parameter E2 has periodic changes, obtain the maximum value related to the time series data G1 or the time M under specific conditions x (G1), and the total time B; S393. Based on time M x (G1) and time B, calculate the encouraging progress parameter R9, the calculation formula of the encouraging progress parameter R9 is as follows: The neuroticism sensitivity parameter calculation process includes: S3100. Get the encouraging progress parameter R9; S3101. Based on the encouragement and progress parameter R9, the neurotic sensitivity parameter R10 is calculated. The calculation formula of the neurotic sensitivity parameter R10 is as follows: R10 = 10τ(R9); The happiness parameter calculation process includes: S3110. Obtain the peak value P corresponding to the maximum frequency in the vibration image frequency histogram; S3111. Based on the standard deviation τ and the peak value P, the happiness parameter R11 is calculated. The calculation formula of the happiness parameter R11 is as follows: The happiness parameter calculation process includes: S3120. Calculate information efficiency X and energy consumption Y based on vibration image data; S3121. Calculate the change in information efficiency dU; calculate the change in energy consumption dV; S3122. Based on the information efficiency X, energy consumption Y, the change in information efficiency dU and the change in energy consumption dV, calculate the happiness parameter R12. The calculation formula of the happiness parameter R12 is as follows:

Images