Summary of the invention
In order to solve the above-mentioned technical problem, the invention proposes a seed moods to incorporate method into.The present invention is specifically with such as
What lower technical solution was realized:
One seed mood incorporates method into, and the method in a manner of sub- mood for describing primary emotion vector, comprising:
Quantify the expression degree of primary emotion;The value of primary emotion vector primary emotion vector is a={ a0,a1,a2,a3,
a4,a5,a6,a7, the meaning of each positional value is followed successively by the degree of being surprised, wild with joy degree, angry degree, grieved degree, vigilant journey
It spends, have deep love for degree, hatred degree, panic degree;
The sub- mood of primary emotion is divided using preset three-level sorting algorithm.
Further, quantify the expression degree of primary emotion using numerical value 0-100, wherein 100 expression are the strongest, 0
It expresses the slightest.
Further, its corresponding first dividing value and the second dividing value, each original feelings are set for each primary emotion
Corresponding first dividing value of thread can be different, and the second dividing value can also be different.
Further, the sub- mood for dividing primary emotion using preset three-level sorting algorithm includes:
a0Less than the first dividing value, then its corresponding sub- mood is dispersion/entanglement, a0Greater than the first dividing value and less than second
Dividing value, corresponding sub- mood are surprised, a0Greater than the second dividing value, corresponding sub- mood is to be surprised;
a1Less than the first dividing value, then its corresponding sub- mood is calmness, a1Greater than the first dividing value and less than the second dividing value,
Its corresponding sub- mood is happy;a1Greater than the second dividing value, corresponding sub- mood is wild with joy;
a2Less than the first dividing value, then its corresponding sub- mood is worry, a2Greater than the first dividing value and less than the second dividing value,
Its corresponding sub- mood is indignant, a2Greater than the second dividing value, corresponding sub- mood is indignation;
a3Less than the first dividing value, then its corresponding sub- mood is to ponder, a3Greater than the first dividing value and less than the second dividing value,
Its corresponding sub- mood is sadness, a3Greater than the second dividing value, corresponding sub- mood is grief;
a4Less than the first dividing value, then its corresponding sub- mood is interest, a4Greater than the first dividing value and less than the second dividing value,
Its corresponding sub- mood is early warning, a4Greater than the second dividing value, corresponding sub- mood is vigilant;
a5Less than the first dividing value, then its corresponding sub- mood is to receive, a5Greater than the first dividing value and less than the second dividing value,
Its corresponding sub- mood is to like, a5Greater than the second dividing value, corresponding sub- mood is to have deep love for;
a6Less than the first dividing value, then its corresponding sub- mood is boring, a6Greater than the first dividing value and less than the second dividing value,
Its corresponding sub- mood is to detest, a6Greater than the second dividing value, corresponding sub- mood is hatred;
a7Less than the first dividing value, then its corresponding sub- mood is worry, a7Greater than the first dividing value and less than the second dividing value,
Its corresponding sub- mood is to fear, a7Greater than the second dividing value, corresponding sub- mood is fear.
Further, further includes:
Primary emotion vector is exported in the form of text.
Further, further includes:
The textual result is the union of the corresponding sub- mood of nonzero value in mood vector.
Further, before the expression degree of the quantization primary emotion, further includes:
Acquire primitive physiological signal;
Obtain the corresponding primary emotion vector of primary emotion vector.
The embodiment of the present invention, which is provided, incorporates method into for a seed mood, can be based on sub- emotion expression service method output character
Emotion judgment has good user experience as a result, in order to understand the emotional state of user.
Specific embodiment
In order to enable those skilled in the art to better understand the solution of the present invention, below in conjunction in the embodiment of the present invention
Attached drawing, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described embodiment is only
The embodiment of a part of the invention, instead of all the embodiments.Based on the embodiments of the present invention, ordinary skill people
The model that the present invention protects all should belong in member's every other embodiment obtained without making creative work
It encloses.
It should be noted that description and claims of this specification and term " first " in above-mentioned attached drawing, "
Two " etc. be to be used to distinguish similar objects, without being used to describe a particular order or precedence order.It should be understood that using in this way
Data be interchangeable under appropriate circumstances, so as to the embodiment of the present invention described herein can in addition to illustrating herein or
Sequence other than those of description is implemented.In addition, term " includes " and " having " and their any deformation, it is intended that cover
Cover it is non-exclusive include, for example, the process, method, system, product or equipment for containing a series of steps or units are not necessarily limited to
Step or unit those of is clearly listed, but may include be not clearly listed or for these process, methods, product
Or other step or units that equipment is intrinsic.
The embodiment of the present invention discloses a kind of mood and incorporates model building method into roughly, as shown in Figure 1, which comprises
S1. raw data set is obtained, each element that the initial data is concentrated includes primitive physiological signal data packet
With primary emotion vector.
The primitive physiological signal data packet is by setting upper pendant in testee when testee is under certain mood
It wears acquisition equipment and obtains.In order to enable testee is among certain mood, mood induction can be carried out to testee.
Specifically, the acquisition of the primitive physiological signal includes following the description:
(1) the skin conductivity signal of testee is measured.
(2) electrode slice is affixed on to the left and right forehead and bilateral ear-lobe of testee, to measure the brain wave letter of testee
Number.
(3) pulse signal and breath signal of testee are measured.
Specifically, model is incorporated into order to establish more reasonable mood, inventor is passed through in the embodiment of the present invention for big
The research achievement for measuring user behavior characteristics and emotion expression service is extracted the primary emotion for being best able to expression human body mood, as feelings
Thread degree incorporates the foundation of model, the primary emotion extracted in the embodiment of the present invention into are as follows: surprising, wild with joy, angry, grieved,
It is vigilant, has deep love for, hating, is panic.Obviously, in people's daily life, most of the time can't generate especially strong
Mood, inventor think that the intensity of mood is also another key factor that mood incorporates model into.
Based on the studies above as a result, the meaning of each position of primary emotion vector is successively in the embodiment of the present invention are as follows: be surprised
Off course degree, angry degree, grieved degree, vigilant degree, has deep love for degree, hatred degree, panic degree at wild with joy degree.
S2. it is concentrated from initial data and extracts emotional information, constructed the element concentrated with original amount and train correspondingly
Collect, the element in the training set is the mood signal phasor that mood signal is constituted, the mood signal phasor and the original
Has one-to-one relationship between beginning mood vector.
Specifically, the mood signal is to carry out data processing to primitive physiological signal data packet and obtain, and the present invention is real
It applies example and discloses specific data processing method, including following the description:
(1) emotional information in skin conductance signal is extracted, method particularly includes:
Discrete sampling is carried out to the skin conductance signal, obtains discrete sampling sequence { i, f (i) }.
According to formulaCalculate the fisrt feature parameter of skin conductance signal.WhereinN is discrete sampling number, and t is predetermined constant, is greater than 10 less than 40.
According to formulaCalculate the second feature parameter of skin conductance signal.Wherein p is default
Constant, value between 0 to 1, N be discrete sampling number, wherein Λ (i-1, p, N-1)=2F1(0,-i+1;-N+1;1/p),
In2F1() indicates Gauss hypergeometric progression.
(2) emotional information in brain wave is extracted, method particularly includes: the frequency of occurrences of statistics α wave, β wave and θ wave.
(3) statistical value of pulse signal and breath signal is calculated.
Therefore, the meaning for each position of mood signal phasor that mood signal corresponding with original amount concentration element is constituted
Successively are as follows: fisrt feature parameter, second feature parameter, the α wave frequency of occurrences, the β wave frequency of occurrences, the θ wave frequency of occurrences, pulse rates
Average value, respiratory intensity average value and respiratory rate average value.
S3. according to the default mind of one-to-one relationship training between the mood signal phasor and the primary emotion vector
Through network model.
Specifically, the neural network model is input with mood signal phasor, with the corresponding grouping of mood signal phasor
Number is output.
Specifically, the neural network model includes input layer, neuronal layers and output layer.The neuronal layers include more
A sublayer, the neuron of each sublayer has reference numeral, and the neuronal quantity of each sublayer is equal.In each sublayer
Portion, adjacent neuron interconnect communication, and the neuron connection communication of each neuron and next layer of reference numeral.
The training method of the neural network is disclosed in embodiments of the present invention, as shown in Figure 2, which comprises
S31. it is that the mood signal phasor is grouped according to primary emotion vector, obtains packet data collection.
The identical primary emotion vector of element position where primary emotion vector maximization is summed up as one group, correspondingly,
Its corresponding mood signal phasor is also summed up as one group.
Therefore, eight groups of training datas are obtained in the embodiment of the present invention.Respectively surprised data group, wild with joy data group, anger
Anger data group, vigilant data group, has deep love for data group, hatred data group and panic data group at grieved data group;Correspondingly, this eight
The packet numbering of group is 0-7.
S32. according to the packet data collection training neural network model, by adjusting each nerve in neural network
The weight of first node obtains target nerve network.In the target nerve network, with any feelings of the packet data convergence
Thread signal phasor is that the output that input obtains is consistent with the group result of the packet data convergence.
In order to more accurately obtain a kind of the incorporating into as a result, the embodiment of the present invention further provides for mood finely draws of mood
Return model building method, the mood finely incorporates each grouping of model building method effect packet data convergence, such as Fig. 3 into
It is shown, which comprises
S100. the consistency of be in a bad mood signal phasor in being grouped is calculated, and chooses the minimum mood signal arrow of consistency
Amount is target emotion signal phasor.
Mood signal phasor a in the embodiment of the present inventioniConsistency beWhereinWherein n is the mood signal phasor number in grouping.
S200. the corresponding target neighborhood collection of the target emotion signal phasor is calculated.
Specifically, the element that the target neighborhood is concentrated is target neighborhood.Target neighborhood in the embodiment of the present invention be with
Centered on target emotion signal phasor, be in a bad mood signal arrow in M-1 times of R radius and M times of the formed annular region of R radius
The set of amount, M are the characteristic value of target neighborhood.R radius is related with the consistency of mood signal phasor in the embodiment of the present invention, tool
Body,Wherein q is adjustable coefficient, O (xi) it is mood signal phasor aiConsistency.Each target
The characteristic value of neighborhood is respectively 1,2 ...,
S300. the K target neighborhood most comprising mood signal phasor quantity is selected.
S400. corresponding target emotion signal phasor is obtained in each target neighborhood as sub- classification center.S500. it obtains
The corresponding primary emotion vector in sub- classification center.
Based on the training result of above-mentioned neural network model, the embodiment of the present invention further provides for a kind of automatic progress mood
The method obtained roughly, as shown in Figure 4, which comprises
S101. collected primitive physiological signal is obtained.
S102. data processing is carried out to the primitive physiological signal, obtains its corresponding mood signal phasor.
S103. the corresponding packet numbering of the mood signal phasor is obtained to obtain target according to the neural network model
Grouping.
S104. the corresponding primary emotion vector of target emotion signal phasor in the targeted packets is exported.
Further, the result of implementation of model building method is finely incorporated into based on mood, the embodiment of the present invention further mentions
For a kind of accurate acquisition methods of mood, the method is further implemented on the basis of the output result of step S104, such as Fig. 5 institute
Show, which comprises
S201. it calculates each height in the targeted packets and sorts out center at a distance from the mood signal phasor.
S202. the nearest corresponding primary emotion vector in sub- classification center of distance is exported.
Further, more fine in order to divide, the embodiment of the present invention further discloses a seed mood and incorporates method into,
The method is emotion expression service and behavioural characteristic progress a large amount of summary research discovery primary emotions differences of the inventor according to user
Expression degree can produce different sub- moods, sub- mood has thus been obtained by the way of quantitative classification incorporates method into,
For the method for describing primary emotion vector in a manner of sub- mood, the method is as shown in Figure 6, comprising:
S10. quantify the expression degree of primary emotion.
Specifically, quantify the expression degree of primary emotion in the embodiment of the present invention using numerical value 0-100, wherein 100 tables
Up to the most strongly, 0 expression is the slightest.
S20. the sub- mood of primary emotion is divided using three-level classification.
Specifically, its corresponding first dividing value and the second dividing value, each primary emotion are set for each primary emotion
Corresponding first dividing value can be different, and the second dividing value can also be different.The value of primary emotion vector is a=in the embodiment of the present invention
{a0,a1,a2,a3,a4,a5,a6,a7, the division methods of sub- mood are as follows:
(1)a0Less than the first dividing value, then its corresponding sub- mood is dispersion/entanglement, a0Greater than the first dividing value and less than
Two dividing values, corresponding sub- mood are surprised, a0Greater than the second dividing value, corresponding sub- mood is to be surprised.
(2)a1Less than the first dividing value, then its corresponding sub- mood is calmness, a1Greater than the first dividing value and less than the second boundary
Value, corresponding sub- mood are happy;a1Greater than the second dividing value, corresponding sub- mood is wild with joy.
(3)a2Less than the first dividing value, then its corresponding sub- mood is worry, a2Greater than the first dividing value and less than the second boundary
Value, corresponding sub- mood are indignant, a2Greater than the second dividing value, corresponding sub- mood is indignation.
(4)a3Less than the first dividing value, then its corresponding sub- mood is to ponder, a3Greater than the first dividing value and less than the second boundary
Value, corresponding sub- mood are sadness, a3Greater than the second dividing value, corresponding sub- mood is grief.
(5)a4Less than the first dividing value, then its corresponding sub- mood is interest, a4Greater than the first dividing value and less than the second boundary
Value, corresponding sub- mood are early warning, a4Greater than the second dividing value, corresponding sub- mood is vigilant.
(6)a5Less than the first dividing value, then its corresponding sub- mood is to receive, a5Greater than the first dividing value and less than the second boundary
Value, corresponding sub- mood are to like, a5Greater than the second dividing value, corresponding sub- mood is to have deep love for.
(7)a6Less than the first dividing value, then its corresponding sub- mood is boring, a6Greater than the first dividing value and less than the second boundary
Value, corresponding sub- mood are to detest, a6Greater than the second dividing value, corresponding sub- mood is hatred.
(8)a7Less than the first dividing value, then its corresponding sub- mood is worry, a7Greater than the first dividing value and less than the second boundary
Value, corresponding sub- mood are to fear, a7Greater than the second dividing value, corresponding sub- mood is fear.
Above-mentioned sub- mood, which incorporates method into, can be used for the output result of processing step S104 and step S202.
Method is incorporated into based on above-mentioned sub- mood, primary emotion vector can be exported in the form of text, thus carrying out feelings
Thread can be with the result of output character when obtaining, it is clear that this would be even more beneficial to the understanding of user.The textual result is mood
The union of the corresponding sub- mood of nonzero value in vector.For example, then output is feared if primary emotion vector is { 0,0,0,0,0,0,98 }
Unbearably.If primary emotion vector is { 56,90,0,0,0,0,0,0 }, then export surprised and wild with joy.If primary emotion vector be 0,0,
78,0,0,0,80,60 }, then export indignant, early warning and be vigilant, hate and fear.
Model is slightly incorporated into the embodiment of the invention provides mood and finely finely incorporates the construction method of model into, and is provided
The automatic method for carrying out mood and incorporating into roughly and finely incorporating into, the physiological signal that can be automatically based upon user judge the feelings of user
Not-ready status has more granularities, automatically, the high remarkable advantage of accuracy;Sub- emotion expression service method is devised simultaneously, and can be based on
The emotion judgment of sub- emotion expression service method output character has good user as a result, in order to understand the emotional state of user
Experience.
It should be understood that referenced herein " multiple " refer to two or more."and/or", description association
The incidence relation of object indicates may exist three kinds of relationships, for example, A and/or B, can indicate: individualism A exists simultaneously A
And B, individualism B these three situations.Character "/" typicallys represent the relationship that forward-backward correlation object is a kind of "or".
The serial number of the above embodiments of the invention is only for description, does not represent the advantages or disadvantages of the embodiments.
Those of ordinary skill in the art will appreciate that realizing that all or part of the steps of above-described embodiment can pass through hardware
It completes, relevant hardware can also be instructed to complete by program, the program can store in a kind of computer-readable
In storage medium, storage medium mentioned above can be read-only memory, disk or CD etc..
The foregoing is merely presently preferred embodiments of the present invention, is not intended to limit the invention, it is all in spirit of the invention and
Within principle, any modification, equivalent replacement, improvement and so on be should all be included in the protection scope of the present invention.