Emotion perception method, system and storage medium
By collecting stress data on the emotion-perceived chair and using long-term and short-term memory neural network analysis, the problem that emotion-perceived systems in the prior art cannot accurately perceive human emotions and privacy leakage is solved, and monitoring and emotional regulation of long-term emotional changes is achieved.
Patent Information
- Application Number
- CN202211020906.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-08-24
- Publication Date
- 2024-11-05
- Estimated Expiration
- 2042-08-24
AI Technical Summary
The existing emotion perception system cannot accurately perceive human emotions without invading user privacy, and it is difficult to obtain the human emotional change curve for a long time, cannot form regular emotional perception, and cannot measure the stress level of the human body.
By installing pressure sensors on the emotion-perceived chair to collect pressure distribution data in the human sitting posture, use long and short-term memory neural network to extract high-dimensional features of the pressure distribution change state, transform it into human emotional state combination features, and obtain emotion types and statistics through classifiers Emotional duration, draw a map of the emotion distribution.
It realizes accurate perception of human emotions without infringing on user privacy, which can reflect emotional changes in the human body during different time periods and helps to regulate emotions.
Smart Images

Figure CN115363583B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of emotion perception technology, and in particular to an emotion perception method, system and storage medium. Background Art
[0002] The accelerated development of society has made the competition faced by human beings increasingly fierce. The psychological burden people bear at work is also increasing, and the losses caused by work pressure are also increasing. There are also other pressures in life. Due to pressure from various aspects, many people are overdrawing their health. The depression caused by stress affects people's mental health and physical health. In recent years, these problems have become increasingly prominent. The perception of emotions can help to adjust the state in time.
[0003] Common emotion perception systems are intelligent gender and emotion recognition detection systems and methods based on vision and voice. Image-based emotion and gender recognition modules are used to perform emotion recognition based on face images and gender recognition based on faces; voice-based emotion and gender recognition modules are used to perform emotion recognition based on human voices and gender recognition based on voices; fusion modules are used to match the gender recognition results and fuse the emotion recognition results, and send them to a personalized intelligent voice interaction system; a personalized intelligent voice interaction system can perform voice interaction. Human emotions are perceived by fusing the recognition results of images and voices. However, the fusion recognition of images and languages involves user privacy and is prone to leakage of user personal information.
[0004] In addition, the existing emotion perception system can only collect facial expressions and voices within a specific short period of time to obtain the current emotional state. It is impossible to obtain the human emotional change curve for a long time based on parameters such as human posture, and it is difficult to form regular emotion perception, which is not conducive to the adjustment of human state and cannot measure and score the stress level of the human body. Summary of the invention
[0005] The present invention provides an emotion perception method, system and storage medium, which are used to solve the problem that the prior art is difficult to accurately perceive human emotions in a way that does not infringe on personal privacy, and to accurately perceive human emotions according to changes in human posture.
[0006] The present invention provides an emotion perception method, comprising:
[0007] Obtain pressure distribution data of human body in sitting position;
[0008] Extracting high-dimensional features of pressure distribution change state from the pressure distribution data through a preset long short-term memory neural network;
[0009] Converting the high-dimensional features of the pressure distribution change state into combined features of human emotional state;
[0010] Classify the combined features of human emotional states through a preset classifier to obtain the emotional type;
[0011] The emotional state is determined according to the emotional type, the emotional duration is counted, and an emotional distribution map is drawn.
[0012] According to an emotion perception method provided by the present invention, the step of obtaining pressure distribution data of a human body in a sitting position specifically includes:
[0013] The pressure sensors distributed on the emotion sensing chair are used to collect the detection values of the pressure sensors;
[0014] Data noise reduction is performed on the detection values of the pressure sensor to generate pressure distribution data.
[0015] According to an emotion perception method provided by the present invention, a high-dimensional feature of the pressure distribution change state is extracted from the pressure distribution data through a preset long short-term memory neural network, specifically comprising:
[0016] Using the pressure distribution data as an input vector, the long short-term memory neural network extracts features of the input vector;
[0017] Classification is performed based on the extracted features, and high-dimensional features of the pressure distribution change state are output.
[0018] According to an emotion perception method provided by the present invention, the step of converting the high-dimensional features of the pressure distribution change state into a combined feature of the human emotional state specifically includes:
[0019] The high-dimensional features of the pressure distribution change state include human breathing frequency features and human posture high-dimensional features;
[0020] The human body breathing frequency feature and the human body posture high-dimensional feature are combined to generate a human body emotional state combined feature.
[0021] According to an emotion perception method provided by the present invention, the method of classifying the combined features of human emotional states by a preset classifier to obtain the emotion type specifically includes:
[0022] Determine the emotion weight according to the proportion of emotion states in the combined characteristics of human emotion states;
[0023] Classify according to emotion weight, the highest emotion weight is the current emotion type;
[0024] Emotion types include: stress level state and joy, sadness, horror, surprise, disgust, and neutral emotion types.
[0025] According to an emotion perception method provided by the present invention, the emotion state is determined according to the emotion type, the emotion duration is counted, and an emotion distribution map is drawn, which specifically includes:
[0026] Determine the current emotional state according to the emotion type, and calculate the duration of the current emotional state;
[0027] When the emotional state changes, the duration of the changed emotional state is recalculated;
[0028] Draw an emotion distribution map based on the duration of different types of emotions.
[0029] The present invention also provides an emotion perception system, the system comprising:
[0030] A data acquisition module, used to acquire pressure distribution data of a human body in a sitting position;
[0031] A feature extraction module, used to extract high-dimensional features of pressure distribution change state from the pressure distribution data through a preset long short-term memory neural network;
[0032] A feature conversion module, used for converting the high-dimensional features of the pressure distribution change state into a combined feature of the human emotional state;
[0033] The emotion classification module is used to classify the combined features of human emotion states through a preset classifier to obtain the emotion type;
[0034] The statistical module is used to determine the emotional state according to the emotional type, count the emotional duration, and draw an emotional distribution map.
[0035] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, any of the above-mentioned emotion perception methods is implemented.
[0036] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the emotion perception method as described above is implemented.
[0037] The present invention also provides a computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the computer program implements any of the above-mentioned emotion perception methods.
[0038] The present invention provides an emotion perception method, system and storage medium, which continuously collects pressure distribution data of a human body in a sitting position, extracts high-dimensional features of the pressure distribution change state from the pressure distribution data, and then transforms the high-dimensional features of the pressure distribution change state into combined features of the human body's emotional state. After classification, the emotion type is obtained, the emotion duration is counted, and the current human emotion is determined. It can also reflect the changes in the human body's emotions in different time periods, which is helpful for emotion regulation. BRIEF DESCRIPTION OF THE DRAWINGS
[0039] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0040] Figure 1 This is one of the flow charts of an emotion perception method provided by the present invention;
[0041] Figure 2 This is the second flow chart of an emotion perception method provided by the present invention;
[0042] Figure 3 This is the third flow chart of an emotion perception method provided by the present invention;
[0043] Figure 4 This is a fourth flow chart of an emotion perception method provided by the present invention;
[0044] Figure 5 This is the fifth flow chart of an emotion perception method provided by the present invention;
[0045] Figure 6 This is the sixth flow chart of an emotion perception method provided by the present invention;
[0046] Figure 7 This is a schematic diagram of module connections of an emotion perception system provided by the present invention;
[0047] Figure 8 is a schematic diagram of the structure of an electronic device provided by the present invention;
[0048] Fig. 9 is a schematic diagram of the structure of the emotion perception chair provided by the present invention;
[0049] Fig.10 This is a long short-term memory network architecture diagram provided by the present invention.
[0050] Reference numerals:
[0051] 1: pressure sensor; 110: data acquisition module; 120: feature extraction module; 130: feature transformation module; 140: emotion classification module; 150: statistics module;
[0052] 810: processor; 820: communication interface; 830: memory; 840: communication bus. DETAILED DESCRIPTION
[0053] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0054] Combine the following Figure 1-Figure 6 An emotion perception method of the present invention is described, comprising:
[0055] S100, obtaining pressure distribution data of a human body in a sitting position;
[0056] S200, extracting high-dimensional features of pressure distribution change state from the pressure distribution data through a preset long short-term memory neural network;
[0057] S300, converting the high-dimensional features of the pressure distribution change state into combined features of human emotional state;
[0058] S400, classifying the combined features of human emotional states through a preset classifier to obtain an emotional type;
[0059] S500: determining the emotional state according to the emotion type, counting the emotion duration, and drawing an emotion distribution map.
[0060] In the present invention, multiple pressure sensors 1 are installed on the emotion perception chair to collect pressure values in different areas and obtain pressure distribution data of the human body in a sitting position. By processing the pressure data, it can be converted into the perception of emotion types, the duration of different emotions can be counted, and an emotion distribution map can be drawn, which is convenient for grasping human emotions and helps to positively regulate human emotions.
[0061] Obtain pressure distribution data of the human body in a sitting position, including:
[0062] S101, collecting detection values of the pressure sensor 1 through the pressure sensor 1 distributed on the emotion sensing chair;
[0063] S102, performing data noise reduction on the detection value of the pressure sensor 1 to generate pressure distribution data.
[0064] refer to Fig. 9 In the present invention, an array of pressure sensors 1 are arranged on the backrest, seat and armrests of the human emotion perception chair. After a person sits on the emotion perception chair, different emotions will correspond to different sitting postures, and the pressures generated by the backrest, seat and armrests in different postures will be different. By analyzing the pressure data of the backrest, seat and armrests, the changes in human emotions can be continuously obtained. In a specific example, 16 pressure sensors 1 are set on the backrest, 12 pressure sensors 1 are set on the seat, and 6 pressure sensors 1 are set on the armrests on both sides. When a person sits on the emotion perception chair, the pressure generated by the backrest, seat and armrests is collected by the pressure sensor 1. As the sitting posture of the human body changes, the pressure values generated in different parts are also different. For example: when a person is in an anxious state, the pressure values of the seat and armrests will be higher than the pressure value of the backrest; when a person is in a relaxed state, the pressure values of the seat, armrests and backrest are more evenly distributed.
[0065] After the pressure data of the pressure sensor 1 is collected, noise reduction processing is performed on the pressure data to remove interference data to ensure the accuracy of the collected pressure data, and pressure distribution data is generated based on the noise-reduced data.
[0066] The high-dimensional features of the pressure distribution change state are extracted from the pressure distribution data by using a preset long short-term memory neural network, specifically including:
[0067] S201, using the pressure distribution data as an input vector, and extracting features of the input vector by the long short-term memory neural network;
[0068] S202, classify according to the extracted features, and output high-dimensional features of the pressure distribution change state.
[0069] The present invention uses a long short-term memory network to process input data. LSTM is a type of neural network containing LSTM blocks or other neural networks. In literature or other materials, LSTM blocks may be described as intelligent network units because they can remember values of indefinite lengths of time. There is a gate in the block that can determine whether the input is important enough to be remembered and whether it can be output.
[0070] refer to Fig.10, at the bottom are four S-function units. The leftmost function may become the input of the block depending on the situation. The three on the right will pass through the gate to determine whether the input can be passed into the block. The second one on the left is the input gate. If the output here is close to zero, the value here will be blocked and will not enter the next layer. The third one on the left is the forget gate. When the value generated here is close to zero, the value remembered in the block will be forgotten. The fourth one is the rightmost input and is the output gate, which can determine whether the input in the block memory can be output.
[0071] The pressure distribution data is used as an input vector, and features in the pressure distribution data are extracted. After the features are extracted, high-dimensional features of the pressure distribution change state are classified and output.
[0072] The high-dimensional features of the pressure distribution change state are transformed into the combined features of the human emotional state, including:
[0073] S301, the high-dimensional features of the pressure distribution change state include human breathing frequency features and human posture high-dimensional features;
[0074] S302: Combining the human body breathing frequency feature and the human body posture high-dimensional feature to generate a human body emotional state combined feature.
[0075] The high-dimensional features of the pressure distribution change state are further processed and divided into human breathing frequency features and human posture high-dimensional features;
[0076] The characteristics of human respiratory rate include: respiratory rate and respiratory depth; respiratory rate is the frequency of human breathing per unit time, and the normal respiratory rate of human body is 16-18 times / minute.
[0077] Respiratory depth is a calculation of the fluctuation of human breathing. It reflects a person's emotional state based on the variation rate of respiratory rate, that is, the changes in breathing speed and depth over a period of time.
[0078] When the human body is in an anxious state, the breathing rate will increase accordingly, and when the human body is in a relaxed state, the breathing rate will slow down accordingly. By collecting the human body's breathing rate, it can help analyze the current emotional state of the human body.
[0079] The high-dimensional features of human posture include: analyzing the current sitting posture of the human body based on the pressure data of different parts of the chair back, seat surface and armrests. The human body has different sitting postures in different emotional states. The human body sitting posture information corresponding to the pressure distribution data of different parts is stored in the preset database. The real-time collected pressure distribution data is compared with the data stored in the database through the preset comparison model to determine the current sitting posture of the human body. The current emotional state of the human body is analyzed based on the sitting posture.
[0080] The human body's breathing frequency features and the high-dimensional features of the human body posture are combined to generate a combined feature of the human body's emotional state, so as to realize a combined analysis of the breathing frequency and the human body's sitting posture, comprehensively determine the current human body's emotional state, and improve the accuracy of emotional judgment.
[0081] The combined features of human emotional states are classified by a preset classifier to obtain the emotional types, including:
[0082] S401, determining an emotion weight according to the proportion of the emotion state in the combined characteristics of the human body emotion state;
[0083] S402, classifying according to emotion weights, the emotion with the highest emotion weight is the current emotion type.
[0084] Emotion types include: stress level state and joy, sadness, horror, surprise, disgust, and neutral emotion types.
[0085] In the combined features of human emotional states, there may be preference features for multiple emotional states to coexist in the analysis and judgment results within a time period. Each emotional state has a certain emotional weight, and the emotional state with the highest emotional weight is determined as the emotional state in the current time period. In a specific example, it is judged that the emotional weight of anxiety in the current time period accounts for 70%, the emotional weight of anger accounts for 16%, and the emotional weight of tension accounts for 14%. Since the emotional weight of anxiety accounts for the largest proportion, it is judged that the emotional state of the human body in the current time period is anxiety.
[0086] Determine the emotional state according to the emotional type, count the emotional duration, and draw an emotional distribution map, specifically including:
[0087] S501, determining a current emotional state according to the emotion type, and calculating a duration of the current emotional state;
[0088] S502, when the emotional state changes, recounting the duration of the changed emotional state;
[0089] S503: Draw an emotion distribution map according to the duration of different types of emotions.
[0090] After a person sits on the emotion perception chair, the pressure sensor 1 can continuously collect pressure data, and the duration of a certain emotion of the person can be calculated. Since the pressure sensor 1 collects data in real time, when the emotional state of the person changes, the sitting posture will also be adjusted accordingly. The pressure sensor 1 can quickly capture the pressure changes and thus determine the changes in the person's emotions. The duration of each emotion is counted, and an emotion distribution map is drawn according to the duration of different types of emotions. The distribution and changes of the person's emotional state can be intuitively judged to achieve the perception of human emotions.
[0091] The present invention provides an emotion perception method, which continuously collects pressure distribution data of a human body in a sitting position, extracts high-dimensional features of the pressure distribution change state from the pressure distribution data, and then transforms the high-dimensional features of the pressure distribution change state into combined features of the human body's emotional state. After classification, the emotion type is obtained, the emotion duration is counted, and the current human body's emotion is determined. The method can also reflect the changes in the human body's emotions in different time periods, which is helpful for emotion regulation.
[0092] refer to Figure 7 The present invention also discloses an emotion perception system, the system comprising:
[0093] The data acquisition module 110 is used to acquire pressure distribution data of a human body in a sitting position;
[0094] A feature extraction module 120, used to extract high-dimensional features of pressure distribution change state from the pressure distribution data through a preset long short-term memory neural network;
[0095] A feature conversion module 130, used to convert the high-dimensional features of the pressure distribution change state into a combined feature of the human emotional state;
[0096] The emotion classification module 140 is used to classify the combined features of the human emotional state through a preset classifier to obtain the emotion type;
[0097] The statistical module 150 is used to determine the emotional state according to the emotional type, count the emotional duration, and draw an emotional distribution map.
[0098] The data acquisition module 110 collects the detection value of the pressure sensor 1 through the pressure sensor 1 distributed on the emotion sensing chair;
[0099] The detection value of the pressure sensor 1 is subjected to data noise reduction to generate pressure distribution data.
[0100] By arranging array pressure sensors 1 on the backrest, seat and armrests of the human emotion sensing chair, when a person sits on the emotion sensing chair, different emotions will correspond to different sitting postures, and the pressures generated by the backrest, seat and armrests will be different under different postures. By analyzing the pressure data of the backrest, seat and armrests, the changes in human emotions can be continuously obtained.
[0101] The feature extraction module 120 uses the pressure distribution data as an input vector, and the long short-term memory neural network extracts the features of the input vector;
[0102] Classification is performed based on the extracted features, and high-dimensional features of the pressure distribution change state are output.
[0103] The feature conversion module 130, the high-dimensional features of the pressure distribution change state include the human breathing frequency features and the human posture high-dimensional features;
[0104] The human body breathing frequency feature and the human body posture high-dimensional feature are combined to generate a human body emotional state combined feature.
[0105] The high-dimensional features of the pressure distribution change state are further processed and divided into human breathing frequency features and human posture high-dimensional features;
[0106] The characteristics of human respiratory rate include: respiratory rate and respiratory depth; respiratory rate is the frequency of human breathing per unit time, and the normal respiratory rate of human body is 16-18 times / minute.
[0107] Respiratory depth is a calculation of the fluctuation of human breathing. It reflects a person's emotional state based on the variation rate of respiratory rate, that is, the changes in breathing speed and depth over a period of time.
[0108] When the human body is in an anxious state, the breathing rate will increase accordingly, and when the human body is in a relaxed state, the breathing rate will slow down accordingly. By collecting the human body's breathing rate, it can help analyze the current emotional state of the human body.
[0109] The high-dimensional features of human posture include: analyzing the current sitting posture of the human body based on the pressure data of different parts of the chair back, seat surface and armrests. The human body has different sitting postures in different emotional states. The human body sitting posture information corresponding to the pressure distribution data of different parts is stored in the preset database. The real-time collected pressure distribution data is compared with the data stored in the database through the preset comparison model to determine the current sitting posture of the human body. The current emotional state of the human body is analyzed based on the sitting posture.
[0110] The human body's breathing frequency features and the high-dimensional features of the human body posture are combined to generate a combined feature of the human body's emotional state, so as to realize a combined analysis of the breathing frequency and the human body's sitting posture, comprehensively determine the current human body's emotional state, and improve the accuracy of emotional judgment.
[0111] The emotion classification module 140 determines the emotion weight according to the emotion state proportion in the combined feature of the human emotion state;
[0112] Classify according to emotion weight, the emotion with the highest weight is the current emotion type.
[0113] This makes it possible to determine the human body's emotions in the current time period, making emotional judgments more accurate.
[0114] A statistics module 150 determines the current emotional state according to the emotion type and calculates the duration of the current emotional state;
[0115] When the emotional state changes, the duration of the changed emotional state is recalculated;
[0116] Draw an emotion distribution map based on the duration of different types of emotions.
[0117] By counting the duration of different emotional states and drawing an emotional distribution map, we can more intuitively distinguish the changes in human emotions.
[0118] The present invention provides an emotion perception system, which continuously collects pressure distribution data of a human body in a sitting position, extracts high-dimensional features of the pressure distribution change state from the pressure distribution data, and then transforms the high-dimensional features of the pressure distribution change state into combined features of the human body's emotional state. After classification, the emotion type is obtained, the emotion duration is counted, and the current human emotion is determined. The system can also reflect the changes in the human body's emotions in different time periods, which is helpful for emotion regulation.
[0119] Figure 8 An example of a physical structure diagram of an electronic device is shown in FIG. 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 may call the logic instructions in the memory 830 to execute an emotion perception method, the method comprising: obtaining pressure distribution data of a human body in a sitting position;
[0120] Extracting high-dimensional features of pressure distribution change state from the pressure distribution data through a preset long short-term memory neural network;
[0121] Converting the high-dimensional features of the pressure distribution change state into combined features of human emotional state;
[0122] Classify the combined features of human emotional states through a preset classifier to obtain the emotional type;
[0123] The emotional state is determined according to the emotional type, the emotional duration is counted, and an emotional distribution map is drawn.
[0124] In addition, the logic instructions in the above-mentioned memory 830 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.
[0125] On the other hand, the present invention further provides a computer program product, the computer program product includes a computer program, the computer program can be stored in a non-transitory computer-readable storage medium, when the computer program is executed by a processor, the computer can execute an emotion perception method provided by the above methods, the method comprising: obtaining pressure distribution data of a human body in a sitting position;
[0126] Extracting high-dimensional features of pressure distribution change state from the pressure distribution data through a preset long short-term memory neural network;
[0127] Converting the high-dimensional features of the pressure distribution change state into combined features of human emotional state;
[0128] Classify the combined features of human emotional states through a preset classifier to obtain the emotional type;
[0129] The emotional state is determined according to the emotional type, the emotional duration is counted, and an emotional distribution map is drawn.
[0130] In another aspect, the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, an emotion perception method provided by the above methods is implemented, the method comprising: obtaining pressure distribution data of a human body in a sitting position;
[0131] Extracting high-dimensional features of pressure distribution change state from the pressure distribution data through a preset long short-term memory neural network;
[0132] Converting the high-dimensional features of the pressure distribution change state into combined features of human emotional state;
[0133] Classify the combined features of human emotional states through a preset classifier to obtain the emotional type;
[0134] The emotional state is determined according to the emotional type, the emotional duration is counted, and an emotional distribution map is drawn.
[0135] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.
[0136] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.
[0137] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. An emotion perception method, characterized in that: include: Obtain pressure distribution data of human body in sitting position; Extracting high-dimensional features of pressure distribution change state from the pressure distribution data through a preset long short-term memory neural network; Converting the high-dimensional features of the pressure distribution change state into combined features of human emotional state; Classify the combined features of human emotional states through a preset classifier to obtain the emotional type; Determine the emotional state according to the emotion type, count the emotion duration, and draw an emotion distribution map; Among them, the pressure distribution data is used as an input vector, the features in the pressure distribution data are extracted, and after the features are extracted, the high-dimensional features of the pressure distribution change state are classified and output; The high-dimensional features of the pressure distribution change state are converted into combined features of the human emotional state, specifically including: The high-dimensional features of the pressure distribution change state include human breathing frequency features and human posture high-dimensional features; The human body breathing frequency feature and the human body posture high-dimensional feature are combined to generate a human body emotional state combined feature.
2. The emotion perception method according to claim 1, characterized in that: The obtaining of pressure distribution data of a human body in a sitting position specifically includes: The pressure sensors distributed on the emotion sensing chair are used to collect the detection values of the pressure sensors; Data noise reduction is performed on the detection values of the pressure sensor to generate pressure distribution data.
3. The emotion perception method according to claim 1, characterized in that: The high-dimensional features of the pressure distribution change state are extracted from the pressure distribution data by using a preset long short-term memory neural network, specifically including: Using the pressure distribution data as an input vector, the long short-term memory neural network extracts features of the input vector; Classification is performed based on the extracted features, and high-dimensional features of the pressure distribution change state are output.
4. The emotion perception method according to claim 1, characterized in that: The method of classifying the combined features of human emotional states by a preset classifier to obtain the emotional type specifically includes: Determine the emotion weight according to the proportion of emotion states in the combined characteristics of human emotion states; Classify according to emotion weight, the emotion with the highest weight is the current emotion type.
5. The emotion perception method according to claim 1, characterized in that: Determine the emotional state according to the emotional type, count the emotional duration, and draw an emotional distribution map, specifically including: Determine the current emotional state according to the emotion type, and calculate the duration of the current emotional state; When the emotional state changes, the duration of the changed emotional state is recalculated; Draw an emotion distribution map based on the duration of different types of emotions.
6. An emotion perception system, characterized in that: The system comprises: A data acquisition module, used to acquire pressure distribution data of a human body in a sitting position; A feature extraction module, used to extract high-dimensional features of pressure distribution change state from the pressure distribution data through a preset long short-term memory neural network; A feature conversion module, used for converting the high-dimensional features of the pressure distribution change state into a combined feature of the human emotional state; The emotion classification module is used to classify the combined features of human emotion states through a preset classifier to obtain the emotion type; A statistical module, used to determine the emotional state according to the emotional type, count the emotional duration, and draw an emotional distribution map; Among them, the pressure distribution data is used as an input vector, the features in the pressure distribution data are extracted, and after the features are extracted, the high-dimensional features of the pressure distribution change state are classified and output; The high-dimensional features of the pressure distribution change state are converted into combined features of the human emotional state, specifically including: The high-dimensional features of the pressure distribution change state include human breathing frequency features and human posture high-dimensional features; The human body breathing frequency feature and the human body posture high-dimensional feature are combined to generate a human body emotional state combined feature.
7. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, an emotion perception method as described in any one of claims 1 to 5 is implemented.
8. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, an emotion perception method as described in any one of claims 1 to 5 is implemented.
9. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, an emotion perception method as described in any one of claims 1 to 5 is implemented.
