CN108630299A - Personality analysis method and system, storage medium based on skin resistance feature - Google Patents
Personality analysis method and system, storage medium based on skin resistance feature Download PDFInfo
- Publication number
- CN108630299A CN108630299A CN201810395841.5A CN201810395841A CN108630299A CN 108630299 A CN108630299 A CN 108630299A CN 201810395841 A CN201810395841 A CN 201810395841A CN 108630299 A CN108630299 A CN 108630299A
- Authority
- CN
- China
- Prior art keywords
- user
- skin resistance
- personality
- resistance signal
- emotion
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H20/00—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance
- G16H20/70—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance relating to mental therapies, e.g. psychological therapy or autogenous training
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/14—Vascular patterns
Abstract
The present invention provides a kind of personality analysis method and system, storage medium based on skin resistance feature, and this method includes：Obtain skin resistance signal of the user to be measured under the stimulation of preset emotion stimulus；Determine the Time-domain Statistics feature and frequency domain statistical nature of the skin resistance signal；By in the Time-domain Statistics feature and the frequency domain statistical nature input emotion prediction model trained in advance, the affective characteristics of the user to be measured are obtained；By in affective characteristics input personality prediction model trained in advance, the personality characteristics of the user to be measured is obtained.The present invention is based on skin resistance features to carry out personality analysis, can improve the accuracy of personality analysis.
Description
Technical field
The present invention relates to personality analysis technical fields, and in particular to a kind of personality analysis method based on skin resistance feature
With system, storage medium.
Background technology
Currently, personality analysis is widely used in the fields such as selection of talented people, psychological consultation and occupational counseling.In the prior art
There is the personality analysis method based on social media, since the confidence level of the information shown in social media is not high, causes to analyze
Personality accuracy it is not high.
Invention content
(1) the technical issues of solving
In view of the deficiencies of the prior art, the present invention provides a kind of personality analysis method based on skin resistance feature and it is
System, storage medium, can improve the accuracy of personality analysis.
(2) technical solution
In order to achieve the above object, the present invention is achieved by the following technical programs：
The present invention provides a kind of personality analysis method based on skin resistance feature, including：
Obtain skin resistance signal of the user to be measured under the stimulation of preset emotion stimulus；
Determine the Time-domain Statistics feature and frequency domain statistical nature of the skin resistance signal；
By in the Time-domain Statistics feature and the frequency domain statistical nature input emotion prediction model trained in advance, obtain
The affective characteristics of the user to be measured；
By in affective characteristics input personality prediction model trained in advance, the personality for obtaining the user to be measured is special
Sign.
The present invention also provides a kind of personality analysis systems based on skin resistance feature, including：
Signal acquisition module, for obtaining skin resistance signal of the user to be measured under the stimulation of preset emotion stimulus；
Characteristic determination module, the Time-domain Statistics feature for determining the skin resistance signal and frequency domain statistical nature；
Sentiment analysis module, the feelings for training the Time-domain Statistics feature and frequency domain statistical nature input in advance
Feel in prediction model, obtains the affective characteristics of the user to be measured；
Personality analysis module, for by affective characteristics input personality prediction model trained in advance, obtaining described
The personality characteristics of user to be measured.
The present invention also provides a kind of computer readable storage mediums, are stored thereon with computer program, which is characterized in that
Processor can realize the above method when executing the computer program.
(3) advantageous effect
Personality analysis method and system, storage medium provided in an embodiment of the present invention based on skin resistance feature, it is to be measured
User generates different emotions under the stimulation of emotion stimulus, and then influences the skin resistance signal of user to be measured, Jin Erji
In skin resistance signal, the affective characteristics of user to be measured are predicted, and then go out the personality of user to be measured according to emotional feature analysis
Feature.Since skin resistance signal is truer relative to social media information, the present invention can improve personality analysis
Accuracy.Moreover, the personality characteristics not only strong and weak difference that the present invention analyzes, can compare science analyzes use to be measured
The truth at family.
Description of the drawings
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technology description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of invention for those of ordinary skill in the art without creative efforts, can be with
Obtain other attached drawings according to these attached drawings.
Fig. 1 is the flow diagram of the personality analysis method based on skin resistance feature in one embodiment of the invention；
Fig. 2 is the schematic diagram of emotion stimulus in one embodiment of the invention；
Fig. 3~8 are the signals for being filtered rear signal in one embodiment of the invention to the skin resistance signal of 6 kinds of emotions
Figure；
Fig. 9 is the structure diagram of the personality analysis system based on skin resistance feature in one embodiment of the invention.
Specific implementation mode
In order to make the object, technical scheme and advantages of the embodiment of the invention clearer, below in conjunction with the embodiment of the present invention
In attached drawing, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described embodiment is
A part of the embodiment of the present invention, instead of all the embodiments.Based on the embodiments of the present invention, those of ordinary skill in the art
The every other embodiment obtained without creative efforts, shall fall within the protection scope of the present invention.
In a first aspect, the present invention provides a kind of personality analysis method based on skin resistance feature, as shown in Figure 1, the party
Method includes：
S101, skin resistance signal of the user to be measured under the stimulation of preset emotion stimulus is obtained；
Wherein, emotion stimulus may include visual transmission source, and the visual transmission source is by multiple video clips and sets
It sets the tranquil segment before each video clip to be formed, the multiple video clip can excite a variety of emotions of user.
For example, as shown in Fig. 2, from multiple segments well-chosen editing go out to be suitble to excite 19 of young people's emotion
Vidclip, this primary experiment of 19 segments composition, primary experiment can inspire six kinds of emotions, be followed successively by it is glad, surprised, detest
It dislikes, is sad, indignation, fear.One segment about 3 minutes or so, tranquil segment are to allow user to be measured in excitation feelings every time
Tranquility is in before sense, and the mood back of the body for recording user to be measured is used as baseline.In addition, the excitation of one section of mood generally exists
Below just can be effective, so the skin resistance signal corresponding to latter 50 seconds of the playing process of one section of video clip can be obtained, and
It is not the skin resistance signal obtained corresponding to the entire playing process of one section of video clip.
Certainly, emotion stimulus can also include other stimulus, for example, sonic stimulation source etc..
Before user to be measured watches video clip, it can be the tool that its finger wears upper profession acquisition skin resistance, adopt
Collect and records skin resistance signal.When acquiring skin resistance signal, per second can extract 120 times, it is then per second to obtain 120
A data point.If after a user to be measured watches a video clip, 50*120 data point can be got.
In practical application, can also be normalized before carrying out subsequent processing according to the skin resistance signal of acquisition
The pretreatment operations such as handle, be filtered.
The reason of being normalized be：The individual difference of electrodermal response foundation level is very big, the skin of different people
The horizontal different or even same person of skin electricity all can be different under different time, varying environment.It is different in order to study
The horizontal relationship between emotion of application on human skin electricity, needs the foundation level difference for removing each user's skin electrical signal, that is,
Individual difference can just work out the variation that certain internal characteristics of electrodermal response are generated with emotion difference.
Specifically normalized process includes：By skin resistance letter of the user to be measured under the stimulation of each video clip
Number with corresponding reference signal make it is poor, obtain normalized skin resistance signal；Wherein, the reference signal is the use to be measured
The mean value of skin resistance signal under tranquil segment of the family corresponding to the video clip.
For example, user to be measured is individually subtracted in 6000 data points for the rear 50s for watching first video clip process
The mean value of all skin resistance signals of tranquil segment of the user to be measured before watching first video clip, obtains 6000
Normalized skin resistance signal, just eliminates individual difference in this way.
The reason of being filtered is that extraneous very faint interference generates prodigious interference after amplifier, or even floods
No useful signal.The characteristics of Butterworth filter (Butterworth), is that the frequency response curve of passband is most smooth, and
With flat to greatest extent, do not rise and fall, and it is zero to be then gradually reduced in suppressed frequency band, this makes it be suitable for many occasions.Skin
The useful frequency band of resistance signal, that is, electrodermal response signal is mainly in 0.2Hz hereinafter, its interference signal and electrodermal response letter
Number frequency band be nonoverlapping, therefore Butterworth filter may be used, the skin resistance signal is filtered；
Wherein, the exponent number of the Butterworth filter is 2, cutoff frequency 0.3Hz.High-frequency Interference can be effectively filtered out in this way,
Removal High-frequency Interference is exactly the process of smoothing processing.For example, Fig. 3~8 are the skin using 6 kinds of emotions of Butterworth filter pair
Resistance signal be filtered after signal schematic representation.
Certainly, the case where being flooded completely by noise for signal then intercepts and abandons these nugatory data.
S102, the Time-domain Statistics feature and frequency domain statistical nature for determining the skin resistance signal；
The step can determine the feature of skin resistance signal in terms of time domain and frequency domain two：
Time domain：
The Time-domain Statistics feature include the mean value of the skin resistance signal, intermediate value, standard deviation, minimum value, maximum value,
At least one of difference, minimum value ratio and maximum value ratio of maxima and minima, the single order of the skin resistance signal
The mean value of difference processing result, intermediate value, standard deviation, minimum value, maximum value, the difference of maxima and minima, minimum value ratio and
At least one of maximum value ratio, and/or, mean value, intermediate value, the standard of the second differnce result of the skin resistance signal
At least one of difference, minimum value, maximum value, the difference of maxima and minima, minimum value ratio and maximum value ratio.Namely
Say, Time-domain Statistics feature include some statistical values of skin resistance signal itself, first-order difference result some statistical values and/or
Some statistical values of second differnce result.Wherein, first-order difference embodies the variation tendency and variation speed of signal, if to one
Order difference has taken absolute value, has been ignored as the trend of variation, and only considers to change speed.First-order difference can be used for detecting signal
The extreme point of part, second differnce can be used for detecting the inflection point of signal part, and then convenient for obtaining the spy of skin resistance signal
Sign.
Frequency domain：
The information that can't see in the time domain, such as the frequency change information of signal, after Fourier transform, frequency domain can
To be clearly seen that.Fourier transformation is divided into continuous and discrete two kinds of situations.We to handle discrete numerical value it is necessary to carry out from
Dissipate Fourier transformation (DFT).But the calculation amount of discrete transform is too big, it is difficult to real-time processing problem, therefore we use quick Fu
In leaf transformation so that operation efficiency is greatly improved.Specially：Fast Fourier change is carried out to the skin resistance signal
It changes, and by the mean value of transformation results, intermediate value, standard deviation, minimum value, maximum value, the difference of maxima and minima, minimum value ratio
It is used as the frequency domain statistical nature at least one of maximum value ratio.
S103, the emotion prediction model for training the Time-domain Statistics feature and frequency domain statistical nature input in advance
In, obtain the affective characteristics of the user to be measured；
Emotion prediction model is trained in advance, and training process may include steps of S201~S204：
Skin resistance signal under emotion stimulus stimulation described in S201, the multiple users of acquisition；
Since the data volume that training uses is bigger, the predictive ability for the model that training obtains is more accurate, therefore the number of user
Measure it is The more the better, for example, 100 users.
Above-mentioned steps can be referred to by obtaining the process of skin resistance signal of each user under the stimulation of emotion stimulus
S101, which is not described herein again.
S202, determine in the multiple user each user in the emotion stimulus under the stimulation of each video clip
Skin resistance signal Time-domain Statistics feature and frequency domain statistical nature, and obtain the user in the post-stimulatory feelings of the video clip
Feel feature；
The Time-domain Statistics of skin resistance signal of each user under the stimulation of each video clip are determined in the step
The process of feature and frequency domain statistical nature can refer to above-mentioned steps S102, and details are not described herein again.
For example, an emotion subjectiveness account can be provided for each user, as shown in Fig. 2, each user
After having watched a video clip, the emotional information for having watched this section of video is filled in account, has six in account
Kind mood：It is glad, surprised, detest, be sad, is indignation, frightened, therefore user can be obtained from emotion subjectiveness account and regarded at this
The post-stimulatory affective characteristics of frequency segment.The form that vector may be used in affective characteristics indicates, for example, [0,0,0,1,1,0], contains
Justice is that a user produces sad and indignation mood after finishing watching a video clip.
S203, by each user in the multiple user each video clip stimulation under skin resistance signal time domain
Statistical nature and frequency domain statistical nature are as first training sample, and the emotion by the user under video clip stimulation
Affective tag of the feature as the first training sample of this, according to corresponding first training sample of the multiple user and institute
Affective tag is stated, emotion tranining database is built；
The step is to build the process of emotion tranining database, and a training sample includes a user in a video
The Time-domain Statistics feature and frequency domain statistical nature of skin resistance signal under segment stimulation, there are one the first training sample tools
Affective tag.
For example, visual transmission source includes 19 vidclips, thus a user watched 19 vidclips it
Afterwards, 19 the first training samples are will produce, 19 affective characteristics are also will produce.
A large amount of first training sample caused by a large number of users and corresponding affective tag are formed into emotion and train number
According to library.
S204, it is based on the emotion tranining database, model training is carried out using support vector machines, it is pre- to obtain the emotion
Survey model.
Detailed process may include：
By the first training sample and the format that is required according to libsvm of corresponding affective tag in emotion tranining database
It arranges.Regularization is carried out with svmscale pairs of the first training sample and corresponding affective tag, data is made to fall on [- 1,1]
In range.Using grid.py cross validations select optimal parameter c and g, then svmtrain using obtain optimal parameter c with
G, linear classifier, RBF kernel functions are trained the first training sample in entire emotion tranining database, obtain supporting to
Amount machine model parameter.After having trained .model model files are as a result saved as, have just obtained six graders, that is, six
The prediction model of kind emotion.
Based on above step S201~S204, emotion prediction model can be obtained, and then by the Time-domain Statistics of user to be measured
Feature and frequency domain statistical nature input emotion prediction model, and then obtain the affective characteristics of user to be measured.
In S104, the personality prediction model for training affective characteristics input in advance, the people of the user to be measured is obtained
Lattice feature.
Personality prediction model is trained in advance, and training process may include steps of S301~S303：
S301, the personality characteristics for obtaining each user in the multiple user；
For example, it after having carried out skin resistance experiment to multiple users, can be provided for each user one big
The personality assessment of five Personality tests tests table, and user fills in the information in relation to Personality test in personality assessment's test table, in turn
Go out five kinds of personalities of user for the filled out content analysis of user：OfNeuroticism (Neuroticism), extropism
(Extroversion), open (Openness), agreeableness (Agreeableness), doing one's duty property
(Conscientiousness).Therefore the personality characteristics of user can be determined according to personality assessment's test table that user fills in.Its
In, personality characteristics is similar with affective characteristics, can also be indicated in the form of vector.
S302, multiple affective characteristics of each user in the multiple user under the stimulation of the multiple video clip are made
For second training sample, and using the personality characteristics of the user as the personality label of the second training sample of this, according to institute
Corresponding second training sample of multiple users and the personality label are stated, personality tranining database is built；
For example, visual transmission source includes 19 vidclips, thus a user watched 19 vidclips it
Afterwards, 19 affective characteristics are will produce, this 19 affective characteristics form second training sample.One the second training sample and one
Personal case marker label are corresponding.
S303, it is based on the personality tranining database, model training is carried out by the way of deep learning, obtains the people
Lattice test model.
Detailed process may include：
The emotion tranining database is divided into five word banks, a word bank corresponds to a kind of personality.It will be in each word bank
Second training sample is according to 8：2 ratio cut partition training set and verification collect, and multilayer sense is trained using the libraries keras based on theano
Machine model is known, including four layers of full articulamentum and one layer sigmoid layers.It is trained training sample as input data, training
Process includes preceding to training and backward training.Forward direction training process is unsupervised learning from bottom to top, i.e., since bottom, one
The past top layer of one layer of layer is trained, in training process, after training study obtains the (n-1)th layer parameter, using n-1 layers of output as n-th
The input of layer, training n-th layer, thus respectively obtains the parameter of each layer；Backward training process is top-down supervised learning, i.e.,
The top-down transmission of training error, is finely adjusted parameter.After the completion of training, five model.pkl model files are obtained, respectively
Corresponding five kinds of personalities.
Based on S301~S303, personality prediction model can be obtained, by the affective characteristics input personality prediction of user to be measured
In model, the value of five kinds of personalities of user to be measured can be obtained, it is user's to be measured that the value of this five kinds of personalities, which forms a vector,
Personality characteristics.
Personality analysis method provided by the invention, user to be measured generate different emotions under the stimulation of emotion stimulus,
And then the skin resistance signal of user to be measured is influenced, and then it is based on skin resistance signal, the affective characteristics of user to be measured are predicted,
And then go out the personality characteristics of user to be measured according to emotional feature analysis.Due to skin resistance signal relative to social media information more
Add really, therefore the present invention can improve the accuracy of personality analysis.Moreover, the personality characteristics that the present invention analyzes is not only strong
Weak difference can compare the truth for analyzing user to be measured of science.
The present invention also provides a kind of personality analysis systems based on skin resistance feature, as shown in figure 9, the system includes：
Signal acquisition module, for obtaining skin resistance signal of the user to be measured under the stimulation of preset emotion stimulus；
Characteristic determination module, the Time-domain Statistics feature for determining the skin resistance signal and frequency domain statistical nature；
Sentiment analysis module, the feelings for training the Time-domain Statistics feature and frequency domain statistical nature input in advance
Feel in prediction model, obtains the affective characteristics of the user to be measured；
Personality analysis module, for by affective characteristics input personality prediction model trained in advance, obtaining described
The personality characteristics of user to be measured.
In some embodiments, the emotion stimulus includes visual transmission source, and the visual transmission source is by multiple videos
Segment and the tranquil segment being arranged before each video clip are formed, and the multiple video clip can excite user's
A variety of emotions.
In some embodiments, the system also includes：
Normalize module, Time-domain Statistics feature and frequency domain for determining the skin resistance signal in characteristic determination module
Before statistical nature, skin resistance signal of the user to be measured under the stimulation of each video clip is believed with corresponding benchmark
Number work is poor, obtains normalized skin resistance signal；Wherein, the reference signal is the user to be measured in the video clip institute
The mean value of skin resistance signal under corresponding calmness segment.
In some embodiments, the system also includes：
First model construction module is specifically included for building emotion prediction model：Emotion described in multiple users is obtained to pierce
Skin resistance signal under stimulus stimulation；Determine in the multiple user each user in the emotion stimulus each regard
The Time-domain Statistics feature and frequency domain statistical nature of skin resistance signal under the stimulation of frequency segment, and the user is obtained in the piece of video
The post-stimulatory affective characteristics of section；By skin resistance signal of each user in the multiple user under the stimulation of each video clip
Time-domain Statistics feature and frequency domain statistical nature as first training sample, and by the user the video clip stimulation under
Affective tag of the affective characteristics as the first training sample of this, according to the corresponding first training sample of the multiple user
Originally with the affective tag, emotion tranining database is built；Based on the emotion tranining database, carried out using support vector machines
Model training obtains the emotion prediction model.
In some embodiments, the system also includes：
Second model construction model is specifically included for building the personality prediction model：Obtain the multiple user
In each user personality characteristics；By multiple feelings of each user in the multiple user under the stimulation of the multiple video clip
Feature is felt as second training sample, and using the personality characteristics of the user as the personality mark of the second training sample of this
Label build personality tranining database according to corresponding second training sample of the multiple user and the personality label；Base
In the personality tranining database, model training is carried out by the way of deep learning, obtains the Personality test model.
In some embodiments, the system also includes：
Filter module, Time-domain Statistics feature and frequency for determining the skin resistance signal in the characteristic determination module
Before the statistical nature of domain, the skin resistance signal is filtered using Butterworth filter；Wherein, the Bart
The exponent number of Butterworth filter is 2, cutoff frequency 0.3Hz.
In some embodiments, the Time-domain Statistics feature includes mean value, intermediate value, the standard of the skin resistance signal
At least one of difference, minimum value, maximum value, the difference of maxima and minima, minimum value ratio and maximum value ratio, the skin
The mean value of the first-order difference handling result of skin resistance signal, intermediate value, standard deviation, minimum value, maximum value, maxima and minima it
At least one of difference, minimum value ratio and maximum value ratio, and/or, the second differnce result of the skin resistance signal
In mean value, intermediate value, standard deviation, minimum value, maximum value, the difference of maxima and minima, minimum value ratio and maximum value ratio
It is at least one.
In some embodiments, characteristic determination module is specifically used for：Fast Fourier is carried out to the skin resistance signal
Transformation, and by the mean value of transformation results, intermediate value, standard deviation, minimum value, maximum value, the difference of maxima and minima, minimum value ratio
At least one of rate and maximum value ratio are used as the frequency domain statistical nature.
It will be appreciated that personality analysis system provided by the invention is corresponding with personality analysis method provided by the invention,
The part such as its explanation in relation to content, citing and advantageous effect can refer to the corresponding portion in personality analysis method, herein not
It repeats again.
The present invention also provides a kind of computer readable storage mediums, are stored thereon with computer program, are executed in processor
Above-mentioned personality analysis method can be realized when the computer program.
It should be noted that herein, relational terms such as first and second and the like are used merely to a reality
Body or operation are distinguished with another entity or operation, are deposited without necessarily requiring or implying between these entities or operation
In any actual relationship or order or sequence.Moreover, the terms "include", "comprise" or its any other variant are intended to
Non-exclusive inclusion, so that the process, method, article or equipment including a series of elements is not only wanted including those
Element, but also include other elements that are not explicitly listed, or further include for this process, method, article or equipment
Intrinsic element.In the absence of more restrictions, the element limited by sentence "including a ...", it is not excluded that
There is also other identical elements in process, method, article or equipment including the element.
The above embodiments are merely illustrative of the technical solutions of the present invention, rather than its limitations；Although with reference to the foregoing embodiments
Invention is explained in detail, it will be understood by those of ordinary skill in the art that：It still can be to aforementioned each implementation
Technical solution recorded in example is modified or equivalent replacement of some of the technical features；And these modification or
It replaces, the spirit and scope for various embodiments of the present invention technical solution that it does not separate the essence of the corresponding technical solution.
Claims (10)
1. a kind of personality analysis method based on skin resistance feature, which is characterized in that including：
Obtain skin resistance signal of the user to be measured under the stimulation of preset emotion stimulus；
Determine the Time-domain Statistics feature and frequency domain statistical nature of the skin resistance signal；
By in the Time-domain Statistics feature and the frequency domain statistical nature input emotion prediction model trained in advance, obtain described
The affective characteristics of user to be measured；
By in affective characteristics input personality prediction model trained in advance, the personality characteristics of the user to be measured is obtained.
2. described to regard according to the method described in claim 1, it is characterized in that, the emotion stimulus includes visual transmission source
Frequency stimulus is made of multiple video clips and the tranquil segment being arranged before each video clip, the multiple video
Segment can excite a variety of emotions of user.
3. according to the method described in claim 2, it is characterized in that, the Time-domain Statistics of the determination skin resistance signal are special
It seeks peace before frequency domain statistical nature, the method further includes：
Skin resistance signal of the user to be measured under the stimulation of each video clip and corresponding reference signal work is poor, it obtains
To normalized skin resistance signal；Wherein, the reference signal is that the user to be measured is flat corresponding to the video clip
The mean value of skin resistance signal under stationary plate section.
4. according to the method described in claim 2, it is characterized in that, the building process of the emotion prediction model includes：
Obtain the skin resistance signal under emotion stimulus stimulation described in multiple users；
Determine skin pricktest of each user in the emotion stimulus under the stimulation of each video clip in the multiple user
The Time-domain Statistics feature and frequency domain statistical nature of signal are hindered, and obtains the user in the post-stimulatory affective characteristics of the video clip；
By the Time-domain Statistics feature of skin resistance signal of each user in the multiple user under the stimulation of each video clip
With frequency domain statistical nature as first training sample, and using the user the video clip stimulation under affective characteristics as
The affective tag of the first training sample of this, according to corresponding first training sample of the multiple user and the emotion mark
Label build emotion tranining database；
Based on the emotion tranining database, model training is carried out using support vector machines, obtains the emotion prediction model.
5. according to the method described in claim 4, it is characterized in that, the building process of the personality prediction model includes：
Obtain the personality characteristics of each user in the multiple user；
Using multiple affective characteristics of each user in the multiple user under the stimulation of the multiple video clip as one article the
Two training samples, and using the personality characteristics of the user as the personality label of the second training sample of this, according to the multiple use
Corresponding second training sample in family and the personality label build personality tranining database；
Based on the personality tranining database, model training is carried out by the way of deep learning, obtains the Personality test mould
Type.
6. according to Claims 1 to 5 any one of them method, which is characterized in that the determination skin resistance signal
Before Time-domain Statistics feature and frequency domain statistical nature, the method further includes：Using Butterworth filter to the skin pricktest
Resistance signal is filtered；Wherein, the exponent number of the Butterworth filter is 2, cutoff frequency 0.3Hz.
7. according to Claims 1 to 5 any one of them method, which is characterized in that the Time-domain Statistics feature includes the skin
The mean value of skin resistance signal, intermediate value, standard deviation, minimum value, maximum value, the difference of maxima and minima, minimum value ratio and most
At least one of big value ratio, the mean value of the first-order difference handling result of the skin resistance signal, intermediate value, standard deviation, most
At least one of small value, maximum value, the difference of maxima and minima, minimum value ratio and maximum value ratio, and/or, it is described
The mean value of the second differnce result of skin resistance signal, intermediate value, standard deviation, minimum value, maximum value, maxima and minima it
At least one of difference, minimum value ratio and maximum value ratio.
8. according to Claims 1 to 5 any one of them method, which is characterized in that determine the frequency domain of the skin resistance signal
Statistical nature, including：
Fast Fourier Transform (FFT) is carried out to the skin resistance signal, and by the mean value of transformation results, intermediate value, standard deviation, minimum
At least one of value, maximum value, the difference of maxima and minima, minimum value ratio and maximum value ratio are united as the frequency domain
Count feature.
9. a kind of personality analysis system based on skin resistance feature, which is characterized in that including：
Signal acquisition module, for obtaining skin resistance signal of the user to be measured under the stimulation of preset emotion stimulus；
Characteristic determination module, the Time-domain Statistics feature for determining the skin resistance signal and frequency domain statistical nature；
Sentiment analysis module, for the Time-domain Statistics feature and the frequency domain statistical nature input emotion trained in advance is pre-
It surveys in model, obtains the affective characteristics of the user to be measured；
Personality analysis module, for by affective characteristics input personality prediction model trained in advance, obtaining described to be measured
The personality characteristics of user.
10. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that execute institute in processor
Claim 1~7 any one of them method can be realized when stating computer program.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810395841.5A CN108630299A (en) | 2018-04-27 | 2018-04-27 | Personality analysis method and system, storage medium based on skin resistance feature |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810395841.5A CN108630299A (en) | 2018-04-27 | 2018-04-27 | Personality analysis method and system, storage medium based on skin resistance feature |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108630299A true CN108630299A (en) | 2018-10-09 |
Family
ID=63694986
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810395841.5A Pending CN108630299A (en) | 2018-04-27 | 2018-04-27 | Personality analysis method and system, storage medium based on skin resistance feature |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108630299A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109498039A (en) * | 2018-12-25 | 2019-03-22 | 北京心法科技有限公司 | Personality assessment's method and device |
| CN109672930A (en) * | 2018-12-25 | 2019-04-23 | 北京心法科技有限公司 | Personality association type emotional arousal method and apparatus |
| CN109683840A (en) * | 2018-12-25 | 2019-04-26 | 北京心法科技有限公司 | The exciting method and device of plyability cognitive information processing mechanism |
| CN109697413A (en) * | 2018-12-13 | 2019-04-30 | 合肥工业大学 | Personality analysis method, system and storage medium based on head pose |
| CN109711291A (en) * | 2018-12-13 | 2019-05-03 | 合肥工业大学 | Personality prediction technique based on eye gaze thermodynamic chart |
| CN110638472A (en) * | 2019-09-27 | 2020-01-03 | 新华网股份有限公司 | Emotion recognition method and device, electronic equipment and computer readable storage medium |
| CN111537056A (en) * | 2020-07-08 | 2020-08-14 | 浙江浙能天然气运行有限公司 | Pipeline along-line third-party construction dynamic early warning method based on SVM and time-frequency domain characteristics |
| CN114650857A (en) * | 2019-11-13 | 2022-06-21 | 日本烟草国际股份有限公司 | Inhaler with mental stress monitoring |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2012144967A1 (en) * | 2011-04-19 | 2012-10-26 | Gryn Vladyslav Kostyantynovych | Method for individual rehabilitation through the creation of a positive emotional background |
| CN104055529A (en) * | 2014-06-19 | 2014-09-24 | 西南大学 | Method for calculating emotional electrocardiosignal scaling exponent |
| CN106650621A (en) * | 2016-11-18 | 2017-05-10 | 广东技术师范学院 | Deep learning-based emotion recognition method and system |
| CN107348962A (en) * | 2017-06-01 | 2017-11-17 | 清华大学 | A kind of personal traits measuring method and equipment based on brain-computer interface technology |
| US20180014739A1 (en) * | 2016-07-13 | 2018-01-18 | Sentio Solutions, Inc. | Unobtrusive emotion recognition system |
| CN107918487A (en) * | 2017-10-20 | 2018-04-17 | 南京邮电大学 | A kind of method that Chinese emotion word is identified based on skin electrical signal |
| 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 |
-
2018
- 2018-04-27 CN CN201810395841.5A patent/CN108630299A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2012144967A1 (en) * | 2011-04-19 | 2012-10-26 | Gryn Vladyslav Kostyantynovych | Method for individual rehabilitation through the creation of a positive emotional background |
| CN104055529A (en) * | 2014-06-19 | 2014-09-24 | 西南大学 | Method for calculating emotional electrocardiosignal scaling exponent |
| US20180014739A1 (en) * | 2016-07-13 | 2018-01-18 | Sentio Solutions, Inc. | Unobtrusive emotion recognition system |
| CN106650621A (en) * | 2016-11-18 | 2017-05-10 | 广东技术师范学院 | Deep learning-based emotion recognition method and system |
| CN107348962A (en) * | 2017-06-01 | 2017-11-17 | 清华大学 | A kind of personal traits measuring method and equipment based on brain-computer interface technology |
| CN107918487A (en) * | 2017-10-20 | 2018-04-17 | 南京邮电大学 | A kind of method that Chinese emotion word is identified based on skin electrical signal |
| 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 |
Non-Patent Citations (2)
| Title |
|---|
| FURKAN GÜRPINAR ETAL.: "Multimodal fusion of audio, scene, and face features for first impression estimation", 《2016 23RD INTERNATIONAL CONFERENCE ON PATTERN RECOGNITION (ICPR)》 * |
| RAMANATHAN SUBRAMANIAN ETAL.: "ASCERTAIN: Emotion and Personality Recognition Using Commercial Sensors", 《IEEE TRANSACTIONS ON AFFECTIVE COMPUTING 》 * |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109697413A (en) * | 2018-12-13 | 2019-04-30 | 合肥工业大学 | Personality analysis method, system and storage medium based on head pose |
| CN109711291A (en) * | 2018-12-13 | 2019-05-03 | 合肥工业大学 | Personality prediction technique based on eye gaze thermodynamic chart |
| CN109498039A (en) * | 2018-12-25 | 2019-03-22 | 北京心法科技有限公司 | Personality assessment's method and device |
| CN109672930A (en) * | 2018-12-25 | 2019-04-23 | 北京心法科技有限公司 | Personality association type emotional arousal method and apparatus |
| CN109683840A (en) * | 2018-12-25 | 2019-04-26 | 北京心法科技有限公司 | The exciting method and device of plyability cognitive information processing mechanism |
| CN110638472A (en) * | 2019-09-27 | 2020-01-03 | 新华网股份有限公司 | Emotion recognition method and device, electronic equipment and computer readable storage medium |
| CN110638472B (en) * | 2019-09-27 | 2022-07-05 | 新华网股份有限公司 | Emotion recognition method and device, electronic equipment and computer readable storage medium |
| CN114650857A (en) * | 2019-11-13 | 2022-06-21 | 日本烟草国际股份有限公司 | Inhaler with mental stress monitoring |
| CN111537056A (en) * | 2020-07-08 | 2020-08-14 | 浙江浙能天然气运行有限公司 | Pipeline along-line third-party construction dynamic early warning method based on SVM and time-frequency domain characteristics |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108630299A (en) | Personality analysis method and system, storage medium based on skin resistance feature | |
| Norman-Haignere et al. | Neural responses to natural and model-matched stimuli reveal distinct computations in primary and nonprimary auditory cortex | |
| US20180276540A1 (en) | Modeling of the latent embedding of music using deep neural network | |
| Chen et al. | The AMG1608 dataset for music emotion recognition | |
| CN108523906A (en) | Personality analysis method and system, storage medium based on pulse characteristics | |
| Khare et al. | Classification of emotions from EEG signals using time‐order representation based on the S‐transform and convolutional neural network | |
| Bailes et al. | Comparative time series analysis of perceptual responses to electroacoustic music | |
| CN103489445B (en) | A kind of method and device identifying voice in audio frequency | |
| Khan et al. | A novel audio forensic data-set for digital multimedia forensics | |
| KR20060110988A (en) | Method for classifying a music genre and recognizing a musical instrument signal using bayes decision rule | |
| Wang et al. | Shallow and deep feature fusion for digital audio tampering detection | |
| Cai | Fault diagnosis of rolling bearing based on empirical mode decomposition and higher order statistics | |
| Yang et al. | A fault diagnosis approach for roller bearing based on improved intrinsic timescale decomposition de-noising and kriging-variable predictive model-based class discriminate | |
| Mahajan | Emotion recognition via EEG using neural network classifier | |
| Rowe et al. | Acoustic auto-encoders for biodiversity assessment | |
| To et al. | A general rule for sensory cue summation: evidence from photographic, musical, phonetic and cross-modal stimuli | |
| Long et al. | Denoising of seismic signals based on empirical mode decomposition-wavelet thresholding | |
| Wataraka Gamage et al. | Speech-based continuous emotion prediction by learning perception responses related to salient events: A study based on vocal affect bursts and cross-cultural affect in AVEC 2018 | |
| Verma et al. | Classification and mapping of sound sources in local urban streets through AudioSet data and Bayesian optimized Neural Networks | |
| Fan et al. | Compound fault diagnosis of rolling element bearings using multipoint sparsity–multipoint optimal minimum entropy deconvolution adjustment and adaptive resonance-based signal sparse decomposition | |
| CN115273904A (en) | Angry emotion recognition method and device based on multi-feature fusion | |
| Lei et al. | Epileptic seizure detection in EEG signals using discriminative Stein kernel-based sparse representation | |
| Huang et al. | Order-statistic filtering Fourier decomposition and its application to rolling bearing fault diagnosis | |
| Heravi et al. | Structural health monitoring by probability density function of autoregressive-based damage features and fast distance correlation method | |
| Lee et al. | A single microphone noise reduction algorithm based on the detection and reconstruction of spectro-temporal features |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181009 |