CN109472290B - Emotion fluctuation model analysis method based on finite-state machine - Google Patents
Emotion fluctuation model analysis method based on finite-state machine Download PDFInfo
- Publication number
- CN109472290B CN109472290B CN201811181916.6A CN201811181916A CN109472290B CN 109472290 B CN109472290 B CN 109472290B CN 201811181916 A CN201811181916 A CN 201811181916A CN 109472290 B CN109472290 B CN 109472290B
- Authority
- CN
- China
- Prior art keywords
- emotion
- state
- emotional
- finite
- fluctuation
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
The invention discloses a finite-state-machine-based emotion fluctuation model analysis method, which comprises the following steps: s1, collecting physiological data by using a physical sign collecting device, and extracting characteristics; s2, carrying out emotion classification and identification on the characteristics by using a support vector machine classifier to obtain the emotion in each time period; s3, describing an emotional state transfer process, and constructing and estimating a state sequence; s4, conducting psychological emotion questionnaire survey, and calculating to obtain an emotional state sequence set; s5, performing PAD emotion scale questionnaire survey, performing PAD model classification training on results to obtain emotion samples, labeling emotion states, and performing supervised learning training; and S6, the emotion fluctuation and the emotion quality are predicted through the emotion state sequence set, the model analysis result and the supervised learning training result. The method can be matched with the existing method, realizes the fusion of subjective and objective aspects, and breaks through the relation between single emotion physiological signals and emotions and the research limitation of only considering objective factors.
Description
Technical Field
The invention relates to an analysis method of an emotion fluctuation model, in particular to an emotion fluctuation model analysis and emotion prediction method based on a finite-state machine, and belongs to the field of emotion calculation in artificial intelligence.
Background
The PAD three-dimensional emotion model is a common model for identifying the emotion state by using a dimension space, and has wide application in the fields of audio and video speech synthesis, emotion calculation and the like. The three-dimensional emotion model for PAD was first proposed by Mehrabian, equal to 1974, and includes three dimensions of pleasure, activation, and dominance. The PAD three-dimensional emotion model mainly has the following characteristics: in the PAD model, each emotion uniquely corresponds to a PAD space coordinate position. After the PAD parameters are normalized, the emotion can be identified by a unique three-dimensional coordinate, and the evaluation with high confidence is realized. In the PAD model, the PAD parameter coordinates are determined through a group of standard emotion scales, and the independence among PAD dimensions can distinguish text emotions in different emotion dimensions more easily.
In the existing research, the following three methods are mainly used for realizing the recognition of emotional states: 1. unsupervised learning method. A Deep Learning Network (DLN), which learns training samples without concept labels (classes) to discover structural knowledge in a set of training samples, is typically unsupervised learning. By adopting a hierarchical structure similar to a neural network, the characteristic representation of the sample in the original space is transformed to a new characteristic space through layer-by-layer characteristic transformation, so that the classification or the prediction is easier. Compared with a method for constructing features by artificial rules, deep learning utilizes big data to learn the features, and can depict rich intrinsic information of the data. 2. And (4) a supervised learning method. Training samples with conceptual labels (classes) are learned to make label (class) predictions as possible for data outside the training sample set. Common supervised learning methods include support vector machines, neural networks, decision trees, bayesian networks, K-nearest neighbors, and hidden markov models. 3. Semi-supervised learning method. The learning method realizes the combination of supervised learning and unsupervised learning, and considers how to train and classify by using a small amount of labeled samples and a large amount of unlabeled samples. Specifically, emotion calculation is a calculation model established by constructing emotional states, which analyzes behavioral and physiological signals and establishes emotional interactions between humans based on the measured emotional states. One of the most important links is "Emotion Recognition", i.e., the prediction and estimation of the corresponding emotional state through the behavior and physiological response of the user.
Generally, emotion recognition research based on physiological signals is numerous at present, but is influenced by various factors such as stimulation acquisition, emotion induction categories, acquisition equipment, feature extraction methods, different dimensionality reduction and classification algorithms and the like, so that the difference of recognition accuracy among the researches is large, and the researches are difficult to compare and aim at use.
A Finite State Machine (FSM) is a computational model abstracted for studying the computational process and certain language classes of finite states, and comprises the following parts: a finite set of states describing different states in the system; an input set for representing different input information accepted by the system; a state transition rule set that expresses rules for a system to transition from one state to another state upon receiving different inputs. The finite state machine mainly includes two types, namely a finite receiver and a finite converter. Finite state machines are a mathematical model with discrete input and output systems, while the correspondence between emotions and external stimuli is also discrete, and the transition of emotional states is also a transition from one particular state to another in a limited emotional state space. The activation of the emotion of an individual based on the FSM emotional state transition is not only related to external input stimulation, but also related to the character and the psychology of the current human body. This inherent property is consistent with the functional description of a finite state machine.
In conclusion, how to provide a method for analyzing and predicting emotions based on an emotion fluctuation model of a finite-state machine on the basis of the prior art effectively realizes the analysis of the emotion fluctuation model and completes emotion prediction, and becomes a new research direction for technicians in the industry.
Disclosure of Invention
In view of the above defects in the prior art, the invention provides a finite-state-machine-based emotion fluctuation model analysis method, which comprises the following steps:
s1, collecting physiological data of a person to be tested by using a physical sign collecting device, uploading the obtained physiological data in real time, and extracting characteristics;
s2, carrying out emotion classification and identification on the extracted features by using a support vector machine classifier to obtain emotion in each time period;
s3, describing an emotional state transfer process through a finite state machine, constructing a state sequence, and estimating the state sequence;
s4, carrying out psychological emotion questionnaire survey on the detected person to obtain the psychological emotion age of the detected person, and calculating to obtain an emotional state sequence set;
s5, performing PAD (PAD application data) emotion scale questionnaire survey on the tested personnel, performing PAD model classification training on results obtained by filling in the questionnaire to obtain an emotion sample, labeling emotion states in the emotion sample, and performing supervised learning training on the emotion sample;
and S6, forecasting future emotion fluctuation and emotion quality through the emotion state sequence set and the analysis result of the emotion fluctuation model and the supervised learning training result of the emotion sample.
Preferably, the vital signs acquisition device comprises a galvanic skin sensor and a heart rate sensor.
Preferably, the emotional state transition process in S3 comprises an initiation of an emotional state until a termination of the emotional state.
Preferably, the constructing the state sequence and estimating the state sequence in S3 specifically includes the following steps:
s31, introducing a finite state machine triple of I, S and M, wherein,
i represents inputting an external stimulus (I) 1 ,i 2 ,i 3 ,i 4 ,…,i n ) Is limited to a non-empty set of cells,
s represents the emotional state of the individual (S) 1 ,s 2 ,s 3 ,s 4 ,…,s n ) A finite non-empty set of (a);
s32, dividing the emotional states into seven types, namely happy state, angry state, surprise state, disgust state, fear state, calmness state and sadness state, and subdividing each emotional state into three types, namely slight state, general state and very special state according to the degree;
s33, selecting the external stimulation time length and the self psychological diathesis as an input set, introducing an intensity factor K to the finite state transition rule set, wherein the size of the K is related to the external stimulation intensity and the self psychological diathesis, and the value range of the K is from 0 to 1.
Preferably, the obtaining of the emotional state sequence set through calculation in S4 specifically includes the following steps:
s41, calculating the sentiment quotient y according to a calculation formula, wherein the sentiment quotient = (psychological emotional age/actual age) × 100% t ；
S42, quoting the sentiment y t And applying the obtained psychological emotional age to a joint formula to obtain an emotional fluctuation model
Wherein c is a constant, c>0，α p Indicating the degree of influence, u, of the mood at time p on the current mood t For white noise process, let u t Is a time-dependent variable, u, taking into account the individual's ability to react to strain on itself to various fluctuations t As a function of time, changes in the temperature,
and
the linear correlation is formed;
s43, solving various parameters and condition variances in the emotional fluctuation model, then obtaining another group of emotional state sequences through description of a finite state machine, and summarizing to obtain an emotional state sequence set.
Preferably, the supervised learning training in S5 includes the following steps: and learning the emotion samples by adopting an artificial neural network in a supervised learning method, and then labeling, classifying and predicting data outside the training sample set.
Compared with the prior art, the invention has the advantages that:
the emotion fluctuation model based on the finite-state machine can be matched with the existing emotion recognition method to realize the fusion of subjective and objective aspects, so that the research limitation that the relation between single emotion physiological signals and emotions is broken through and only objective factors are considered is broken through. The invention can make a personal emotion fluctuation model according to the needs so as to prevent and avoid side effects caused by bad emotion. The finite state machine used in the invention can also classify the emotion more finely, thus realizing more accurate description of the subtle emotion.
Meanwhile, the emotional state sequence obtained through the physical sign acquisition equipment and the emotional state sequence analyzed through the emotion fluctuation are effectively combined, so that the fluctuation situation of the emotion is observed more accurately, and the future emotion fluctuation is predicted.
In addition, the invention also provides reference for other related problems in the same field, can be expanded and extended on the basis of the reference, is applied to other emotion fluctuation model analysis and emotion prediction technical schemes in the same field, and has very wide application prospect.
The following detailed description of the embodiments of the present invention is provided in connection with the accompanying drawings for the purpose of facilitating understanding and understanding of the technical solutions of the present invention.
Drawings
FIG. 1 is a schematic diagram of the emotion transfer process in the present invention;
FIG. 2 is a schematic flow chart of the present invention.
Detailed Description
As shown in fig. 1-2, the present invention discloses a finite-state-machine-based emotion fluctuation model analysis method, which comprises the following steps:
s1, collecting physiological data of a detected person by using a sign collecting device, uploading the obtained physiological data in real time, and extracting features. The sign acquisition device comprises a skin electric sensor and a heart rate sensor.
And S2, carrying out emotion classification and identification on the extracted features by using a Support Vector Machine (SVM) classifier to obtain the emotion in each time period.
And S3, describing an emotional state transition process through a finite state machine, constructing a state sequence, and estimating the state sequence. The emotional state transition process includes an initiation of an emotional state until a termination of the emotional state.
The method for constructing the state sequence and estimating the state sequence specifically comprises the following steps:
s31, introducing a finite state machine triple of I, S and M, wherein,
i denotes input of external stimulus (I) 1 ,i 2 ,i 3 ,i 4 ,…,i n ) Is limited to a non-empty set of cells,
s represents an individualEmotional State(s) 1 ,s 2 ,s 3 ,s 4 ,…,s n ) Is limited to a non-empty set of cells,
m is represented as a fuzzy subset on the product space I × S, called fuzzy relation from I to S.
And S32, dividing the emotional states into seven types, namely happy state, angry state, surprise state, disgust state, fear state, calm state and sadness state, and subdividing each emotional state into three types, namely slight state, normal state and extreme state according to degree.
For example, the states of anger, joy, etc. are subdivided into slight anger, general anger and extraordinary anger. S can be represented as (slightly angry, generally angry, very angry, slightly happy, generally happy, very happy) \8230;) a total of 21 states. The change of the emotional state of a person is greatly related to external stimulation, the mind and character of the person, the time length of the stimulation and the like.
S33, selecting the external stimulation time length and the self psychological diathesis as an input set, and introducing an intensity factor K to the finite state transition rule set, wherein the size of the K is related to the external stimulation intensity and the self psychological diathesis, and the value range of the K is from 0 to 1.
Under the condition that specific emotions have been obtained by the SVM classifier, for example, the emotion after feature extraction is happy, then slight happiness is observed when 0< -k < = 0.33. When 0.33< -K < =0.66, it is general happy. When 0.66< -K < =1, it is very happy.
And S4, carrying out psychological emotion questionnaire survey on the tested person to obtain the psychological emotion age of the tested person, and obtaining an emotional state sequence set through calculation.
The method for obtaining the emotional state sequence set through calculation specifically comprises the following steps:
s41, according to the calculation formula, the sentiment quotient = (psychological emotional age/actual age) × 100%, the sentiment quotient is a short term for emotional intelligence, and the fluctuation thereof is positively correlated with the fluctuation of emotion. Calculating the sentiment quotient y t 。
S42, quoting the sentiment y t And applying the obtained psychological emotional age to a joint formula to obtain an emotional fluctuation model
Wherein c is a constant, c>0，α p Indicating the degree of influence, u, of the mood at time p on the current mood t For white noise process, let u t Is a time-dependent variable, u, taking into account the individual's ability to react to strain on itself to various fluctuations t As a function of time, changes in the temperature,
and
and are linearly related.
S43, solving various parameters and condition variances in the emotional fluctuation model, then obtaining another group of emotional state sequences through description of a finite state machine, and summarizing to obtain an emotional state sequence set.
The solution process for the mood swing model is described further below,
y t ＝c+α 1 y t-1 +α 2 y t-2 +…+α p y t-p +u t ， (1)
wherein c is a constant, c>0，α p Indicating the degree to which the mood at time p has an effect on the current mood. U in the equation t For white noise processes, this variable is set to take into account the individual's ability to react to various fluctuations, i.e., a self-regulating control ability. u. u t Is a time-dependent variable, i.e. u t To change over time. One method is to show
And
are linearly related, i.e.:
in view of the above, it can be seen that,
because u is t Is random, and
unlikely to be negative, ξ>0，0<＝β 1 +β 2 +…+β m <1. When these conditions are satisfied, the control unit may,
may be expressed as:
this means that
Is expected to be constant under assumed conditions and can be derived by estimation. (5) The formula shows that the mood fluctuation of the previous period of time can cause the mood fluctuation of the current day to be still large unless other reverse impacts counteract, and similarly, the mood of the current day is calm even and calm unless sudden large impact occurs to cause the mood to be extremely calm
Increasing sharply. In the estimation, we can consider the expression (5) as a form, i.e. assume:
wherein { v t Are independently distributed sequences with zero mean and unit variance. If so:
then there are
This is exactly the formula (5), and this change indicates if u t When equations (6) and (7) are satisfied, equation (5) is satisfied, and the condition just established is desirably not changed.
At the time of estimation, the formula (1) is rewritten into the following form for simplification:
wherein, x' t Is a vector representing all the explanatory variables, which obviously contains the hysteresis value of y, u t The condition of the formula (2) is still satisfied. The first m observations are taken as conditions (t = -m +1, -m +2, \ 8230;, 0). The observed value of T =1,2, \8230;, T was used for estimation. Let Y t The vector representing the observed value obtained within the time t is then
Y t ＝(y t ,y t-1 ,…,y 0 ,…,y -m+1 ,x′ t ,…,x′ 1 ,x′ 0 ,…x′ m+1 )′。
Suppose v t I.i.d N (0, 1), i.e., v t Subject to independent identityDistribution process, then y t|t-1 ～N(x′ t ,β,h t ) Thus, it is possible to
Wherein, δ = (ξ, β) 1 ,…,β m )′,
All the parameters to be estimated are grouped into a vector θ: θ = (β ', δ ') ', so the preceding m log-conditional likelihood functions are:
taking the derivative of equation (12) with respect to θ, which can be done by taking the derivative of equation (10) with respect to θ and then summing, so that:
since the parameter θ can be obtained by the maximum/minimum extremum method in equation (14), the parameters in the mood swing model can be obtained, and the conditional variance can be obtained from equation (10).
And S5, performing PAD (PAD application data) emotion scale questionnaire survey on the tested personnel, performing PAD model classification training on the result obtained by filling the questionnaire to obtain an emotion sample, labeling the emotion state in the emotion sample, and performing supervised learning training on the emotion sample. And learning the emotion samples by adopting a common artificial neural network in a supervised learning method, and then labeling, classifying and predicting data outside the training sample set.
And S6, forecasting future emotion fluctuation and emotion quality through the analysis results of the emotion state sequence set and the emotion fluctuation model and by combining the supervised learning training results of the emotion samples. Through long-time data accumulation and data analysis and comparison, the emotion difference calculated by the emotion state sequence set and the emotion fluctuation model obtained by measurement of the physical sign acquisition equipment is small, and the two modes can be effectively combined to observe the fluctuation situation of emotion more accurately and predict future emotion fluctuation.
The emotion fluctuation model based on the finite-state machine can be matched with the existing emotion recognition method to realize the fusion of subjective and objective aspects, so that the research limitation that the relation between single emotion physiological signals and emotions is broken through and only objective factors are considered is broken through. The invention can make a personal emotion fluctuation model according to the needs so as to prevent and avoid side effects caused by bad emotion. The finite state machine used in the invention can also classify the emotion more finely, thus realizing more accurate description of the subtle emotion.
Meanwhile, the emotional state sequence obtained by the physical sign acquisition equipment (the skin electricity and the heart rate sensor) and the emotional state sequence analyzed by the emotional fluctuation are effectively combined, so that the fluctuation condition of the emotion is observed more accurately, and the future emotional fluctuation is predicted.
In addition, the invention also provides reference for other related problems in the same field, can be expanded and extended on the basis of the reference, is applied to other emotion fluctuation model analysis and emotion prediction technical schemes in the same field, and has very wide application prospect.
It will be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential attributes thereof. The present embodiments are therefore to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein, and any reference signs in the claims are not intended to be construed as limiting the claim concerned.
Furthermore, it should be understood that although the present description refers to embodiments, not every embodiment may contain only a single embodiment, and such description is for clarity only, and those skilled in the art should integrate the description, and the embodiments may be combined as appropriate to form other embodiments understood by those skilled in the art.
Claims (5)
1. A finite state machine-based emotion fluctuation model analysis method is characterized by comprising the following steps: the method comprises the following steps:
s1, collecting physiological data of a person to be tested by using a physical sign collecting device, uploading the obtained physiological data in real time, and extracting characteristics;
s2, carrying out emotion classification and identification on the extracted features by using a support vector machine classifier to obtain emotion in each time period;
s3, describing an emotional state transfer process through a finite state machine, constructing a state sequence, and estimating the state sequence;
s4, carrying out psychological emotion questionnaire survey on the detected person to obtain the psychological emotion age of the detected person, and obtaining an emotional state sequence set through calculation;
s41, calculating the sentiment quotient y according to a calculation formula, wherein the sentiment quotient = (psychological emotional age/actual age) × 100% t ；
S42, quoting the sentiment y t And applying the obtained psychological emotional age to a joint formula to obtain an emotional fluctuation model
Wherein c is a constant, c > 0, α p Indicating the degree of influence, u, of the mood at time p on the current mood t For white noise process, u is set t Is a time-dependent variable, u, taking into account the individual's ability to react to strain on itself to various fluctuations t As a function of time, and is,
and
the linear correlation is formed;
s43, solving various parameters and condition variances in the emotional fluctuation model, then describing by a finite state machine to obtain another group of emotional state sequences, and summarizing to obtain an emotional state sequence set;
s5, PAD emotion scale questionnaire survey is conducted on the tested personnel, the results obtained by questionnaire filling are subjected to PAD model classification training to obtain emotion samples, emotion states in the emotion samples are labeled, and then supervised learning training is conducted on the emotion samples;
and S6, forecasting future emotion fluctuation and emotion quality through the emotion state sequence set and the analysis result of the emotion fluctuation model and the supervised learning training result of the emotion sample.
2. The finite state machine-based mood swing model analysis method of claim 1, characterized in that: the sign acquisition device comprises a skin electric sensor and a heart rate sensor.
3. The finite state machine-based mood swing model analysis method of claim 1, characterized in that: and the emotional state transition process in the S3 comprises the initiation of the emotional state until the termination of the emotional state.
4. The finite state machine-based emotion fluctuation model analysis method of claim 1, wherein the constructing and estimating the state sequence in S3 specifically comprises the following steps:
s31, introducing a finite state machine triple of I, S and M, wherein,
i represents inputting an external stimulus (I) 1 ，i 2 ，i 3 ，i 4 ，…，i n ) Is limited to a non-empty set of (a),
s represents the emotional state of the individual (S) 1 ，s 2 ，s 3 ，s 4 ，…，s n ) Is limited to a non-empty set of cells,
m is expressed as a fuzzy subset on the product space I × S, and is called as a fuzzy relation from I to S;
s32, dividing the emotional states into seven types, namely happy state, angry state, surprise state, disgust state, fear state, calmness state and sadness state, and subdividing each emotional state into three types, namely slight state, normal state and extreme state according to the degree;
s33, selecting the external stimulation time length and the self psychological diathesis as an input set, and introducing an intensity factor K to a finite state transition rule set, wherein the size of the K is related to the external stimulation intensity and the self psychological diathesis, and the value range of the K is from 0 to 1.
5. The finite-state-machine-based emotion fluctuation model analysis method of claim 1, wherein the supervised learning training in S5 comprises the following steps: and learning the emotion samples by adopting an artificial neural network in a supervised learning method, and then labeling, classifying and predicting data outside the training sample set.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811181916.6A CN109472290B (en) | 2018-10-11 | 2018-10-11 | Emotion fluctuation model analysis method based on finite-state machine |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811181916.6A CN109472290B (en) | 2018-10-11 | 2018-10-11 | Emotion fluctuation model analysis method based on finite-state machine |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109472290A CN109472290A (en) | 2019-03-15 |
| CN109472290B true CN109472290B (en) | 2022-10-14 |
Family
ID=65663369
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811181916.6A Active CN109472290B (en) | 2018-10-11 | 2018-10-11 | Emotion fluctuation model analysis method based on finite-state machine |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109472290B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110147432B (en) * | 2019-05-07 | 2023-04-07 | 大连理工大学 | Decision search engine implementation method based on finite state automaton |
| CN111723867A (en) * | 2020-06-22 | 2020-09-29 | 山东大学 | Intelligent evaluation system and method for online game fascination degree |
| CN112263252B (en) * | 2020-09-28 | 2024-05-03 | 贵州大学 | PAD emotion dimension prediction method based on HRV features and three-layer SVR |
| CN112618911B (en) * | 2020-12-31 | 2023-02-03 | 四川音乐学院 | Music feedback adjusting system based on signal processing |
| CN114863636A (en) * | 2022-03-25 | 2022-08-05 | 吉林云帆智能工程有限公司 | Emotion recognition algorithm for rail vehicle driver |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106419936A (en) * | 2016-09-06 | 2017-02-22 | 深圳欧德蒙科技有限公司 | Emotion classifying method and device based on pulse wave time series analysis |
| CN107944473A (en) * | 2017-11-06 | 2018-04-20 | 南京邮电大学 | A kind of physiological signal emotion identification method based on the subjective and objective fusion of multi-categorizer |
| CN107766898A (en) * | 2017-12-08 | 2018-03-06 | 南京邮电大学盐城大数据研究院有限公司 | The three classification mood probabilistic determination methods based on SVM |
-
2018
- 2018-10-11 CN CN201811181916.6A patent/CN109472290B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN109472290A (en) | 2019-03-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109472290B (en) | Emotion fluctuation model analysis method based on finite-state machine | |
| Marechal et al. | Survey on AI-Based Multimodal Methods for Emotion Detection. | |
| Lotfian et al. | Curriculum learning for speech emotion recognition from crowdsourced labels | |
| US11769056B2 (en) | Synthetic data for neural network training using vectors | |
| Vinola et al. | A survey on human emotion recognition approaches, databases and applications | |
| Du et al. | Vision-based fatigue driving recognition method integrating heart rate and facial features | |
| JP5520886B2 (en) | Behavior model learning apparatus, method, and program | |
| Oyedotun et al. | Prototype-incorporated emotional neural network | |
| Amor et al. | Toward sign language handshapes recognition using Myo armband | |
| CN110147822B (en) | Emotion index calculation method based on face action unit detection | |
| Gomes et al. | Establishing the relationship between personality traits and stress in an intelligent environment | |
| Yulita et al. | Multichannel electroencephalography-based emotion recognition using machine learning | |
| Fardian et al. | Thermography for emotion recognition using deep learning in academic settings: A review | |
| Kumar et al. | Analysis of machine learning algorithms for facial expression recognition | |
| Samanta et al. | Emotion sensing from head motion capture | |
| Khorrami | How deep learning can help emotion recognition | |
| KR20210023170A (en) | Deep learning based emotional recognition system and methods using PPG signals | |
| Jadhav et al. | Big five personality prediction using machine learning algorithms | |
| Liliana et al. | The Fuzzy Emotion Recognition Framework Using Semantic-Linguistic Facial Features | |
| Flores-Juarez et al. | Two facial emotion detection based on naive Bayesian classifier | |
| Alamgir et al. | Hybrid multi-modal emotion recognition framework based on InceptionV3DenseNet | |
| Kim et al. | A sensory control system for adjusting group emotion using Bayesian networks and reinforcement learning | |
| Mangla et al. | Employing machine learning for multi-perspective emotional health analysis | |
| Callejas-Cuervo et al. | Architecture of an emotion recognition and video games system to identify personality traits | |
| Rybak et al. | New real-time methods for operator situational awareness retrieval and higher process safety in the control room |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |