CN105469116A - Picture recognition and data extension method for infants based on man-machine interaction - Google Patents
Picture recognition and data extension method for infants based on man-machine interaction Download PDFInfo
- Publication number
- CN105469116A CN105469116A CN201510862736.4A CN201510862736A CN105469116A CN 105469116 A CN105469116 A CN 105469116A CN 201510862736 A CN201510862736 A CN 201510862736A CN 105469116 A CN105469116 A CN 105469116A
- Authority
- CN
- China
- Prior art keywords
- data
- subject image
- category
- man
- machine interaction
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/40—Software arrangements specially adapted for pattern recognition, e.g. user interfaces or toolboxes therefor
- G06F18/41—Interactive pattern learning with a human teacher
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Theoretical Computer Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Human Computer Interaction (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a picture recognition and data extension method for infants based on man-machine interaction. A man-machine interaction platform comprises a server unit and a plurality of client units. The method comprises the following steps: storing initialization data into a server unit, wherein the initialization data is the data of an A type; obtaining an object image through photographing or a network, and uploading the object image to the server unit, wherein the photographed or obtained object image is the data of a B type; respectively calculating the weight of a score obtained in each man-machine interaction process for a certain object image of the B-type data according to the results of multiple man-machine interactions or multiple man-machine interaction processes, determining the final object name of the object image according to the size of the weight, and enabling the object image to be converted into the A-type data from the B-type data. An intelligent robot employing the above method can automatically complete the screening and marking of no-mark data, and converts the no-mark data into mark data, so as to enlarge the data capacity.
Description
Technical field
The child that the present invention relates to a kind of man-machine interactive platform knows drawing method, especially relates to a kind of by carrying out automatic mark to new data to expand child's knowledge figure and the data extending method of content.
Background technology
In recent years, the technology such as computing machine, internet, artificial intelligence achieve breakthrough development, and the terminal devices such as e.g., in hardware device, the computer platform of high-performance, low cost, low-power consumption is rapidly developed, smart mobile phone are used widely; In network service, Wifi, 3G, 4G universal makes people can whenever and wherever possible by wireless network high speed transmission data; In speech recognition, the speech recognition technology of the company such as Google, Microsoft, Baidu exploitation to Chinese or English discrimination up to more than 95%; In image recognition, human detection and face recognition technology are tending towards ripe and widely apply.The research and development developing into Intelligent Service humanoid robot of this this technology and application are laid a good foundation.At present, the application of robot just from the specialized field in past by the normal sphere of life of the order expanding to greatly people, particularly can provide accompany and attend to, the Intelligent Service humanoid robot of the service such as early education just becoming application focus.
But still there is very large room for promotion in existing service type robot in intelligence degree, service content.Such as, to in the image recognition of different objects, need the sample database of a large amount of tape label for training, and the workload gathering these data itself is very huge, need time and the manpower of at substantial, be far from the work that a few peoples just can complete, this is the major reason affecting robot recognition capability, level of intelligence.
Summary of the invention
The object of this invention is to provide a kind of child's knowledge figure based on man-machine interaction and data extending method, there is intelligence, automatic, practical advantage, adopt the intelligent robot of the inventive method automatically can screen data untagged, mark, and be translated into flag data, with expanding data capacity, the intelligent level of hoisting machine people.
For solving in prior art service type robot in the image recognition of different objects, need the sample data gathering a large amount of tape label for training, waste time and energy and affect the technical matters of robot recognition capability and level of intelligence, the invention provides a kind of child's knowledge figure based on man-machine interaction and data extending method, wherein man-machine interactive platform comprises server unit and multiple client unit, server unit is communicated by wired or wireless mode with client unit, and described child's knowledge figure and data extending method comprise the following steps:
One, by initialization data stored in the database of server unit, initialization data comprises subject image and corresponding object names; Initialization data is divided into category-A data also for each subject image of category-A data distributes unique number, described numbering is for identifying this subject image;
Two, by client unit shot object image or from Network Capture subject image, will take or upload onto the server the database of unit from the subject image of Network Capture; Be divided into category-B data also for each subject image of category-B data distributes unique number by shooting or from the subject image of Network Capture, described numbering is for identifying this subject image;
Three, man-machine interaction participant is loaded into the subject image of category-A data in server unit database or category-B data at random by client unit, and answers corresponding object names according to the subject image be loaded into;
Four, for the subject image number of interactive process loading category-A data, answer correctness according to man-machine interaction participant calculates gained achievement score, and records the object names of man-machine interaction participant for each subject image answer of category-B data;
Five, achievement score step 4 obtained and man-machine interaction participant to upload onto the server unit in the lump for the object names that each subject image of category-B data is answered, as the man-machine interaction result that this interactive process returns;
Six, according to step one to the mode of step 5, upload onto the server all results participating in interactive process by client unit unit;
Seven, for the jobbie image of category-B data, according to repeatedly or the result of multiple interactive process calculate the weights of each interactive process gained achievement score respectively, and determine the final object names of this subject image according to the size of weights;
The final object names of certain the category-B data subject image eight, obtained according to step 7, marks the subject image of these category-B data, is category-A data by it from category-B data transformations.
Further, a kind of child based on man-machine interaction of the present invention knows figure and data extending method, wherein, in above-mentioned steps four, the described subject image number for interactive process loading category-A data, answer correctness according to man-machine interaction participant calculates gained achievement score, specifically comprises the following steps:
(1) be loaded into the subject image sum of category-A data by accumulation calculating this interactive process, and be recorded as m;
(2) according to being loaded into the numbering of often opening the subject image of category-A data, the object names answer man-machine interaction participant and the object names of this subject image corresponding stored in a database compare, if consistent, be judged as answering correctly, if inconsistent, be judged as erroneous answers, statistics answers correct number of times, and is recorded as m
1
(3) with answering the subject image total number of correct number of times divided by being loaded into category-A data, using gained divisor as achievement score, and s=m is designated as
1/ m.
Further, a kind of child based on man-machine interaction of the present invention knows figure and data extending method, wherein, in above-mentioned steps seven, described basis repeatedly or the result of multiple interactive process calculate the weights of each interactive process gained achievement score respectively, and the size of foundation weights determines the final object names of this subject image, specifically comprises the following steps:
(1) for a certain subject image of category-B data, the number of times returning man-machine interaction result is set to n, achievement score n interactive process returned is set to set s
i={ s
1, s
2, s
3..., s
n; Man-machine interaction participant n interactive process returned is set to set d for the object names that this subject image is answered
i={ d
1, d
2, d
3..., d
n;
(2) object names answered for this subject image according to the whether identical man-machine interaction participant returned n interactive process of content is distinguished, and obtains the different object names of k kind, be designated as and gather d according to the identical content mode only retained once
j={ d
1, d
2, d
3..., d
k.
(3) according to set d
j={ d
1, d
2, d
3..., d
kin each element, with d
j=d
ifor the achievement score that each element of condition accumulation calculating is corresponding, as the mark weight w that each element is corresponding
i,
w
i＝∑(s
i|d
j＝d
i)
(4) corresponding to each element mark weight w
ibe normalized, as the normalization weights that each element is corresponding
(5) for the result that step (four) obtains, normalization weights corresponding for each element are sorted from big to small and add up successively, until added the threshold value t that accumulative normalization weights sum is greater than setting, wherein 0 < t < 1;
(6) judge according to the number adding accumulative normalization weights in step (five),
If add the number j=1 of accumulative normalization weights:, direct by element d corresponding for maximum normalization weights
jas the final object names to this subject image;
If add the number j=2 of accumulative normalization weights, 3: the method for then being analysed and compared by the meaning of a word judges corresponding several element d
jwhether be synonym, if synonym or the word for comprising justice, then simultaneously by these elements d
jas the final object names of this subject image;
If add the number j=2 of accumulative normalization weights, several element d corresponding when 3
jit is not synonym, or when adding the number j>3 of accumulative normalization weights, the condition that then represents does not meet cannot determine final object names, and by increasing the number of times returning man-machine interaction result, continue to perform step () to step (five), until determine the final object names of this subject image.
Further, a kind of child based on man-machine interaction of the present invention knows figure and data extending method, and wherein, the value of described threshold value t is 0.7.
Further, a kind of child based on man-machine interaction of the present invention knows figure and data extending method, wherein, and a corresponding one or more object names of subject image in described category-A data, and object names multiple subject image corresponding.
Further, a kind of child based on man-machine interaction of the present invention knows figure and data extending method, and wherein, the subject image of described category-A data and category-B data all adopts the image of fixed size, and makes object be positioned at image centre position, and stores with jpg, bmp form.
Further, a kind of child based on man-machine interaction of the present invention knows figure and data extending method, and wherein, the object names of described category-A data adopts one or more Chinese-character words to represent.
A kind of child based on man-machine interaction of the present invention knows figure and data extending method compared with prior art, have the following advantages: the present invention proposes a kind of man-machine interaction knowledge figure and online data extending method, in practical application, server unit is allowed to be communicated by network with the robot as client unit, while making human and computer people carry out man-machine interaction knowledge figure, the new data gathered can be screened, marked, hold with automatic expanding data; The a large amount of new datas obtained can also be used for being trained by machine learning, the level of intelligence of hoisting machine people.This mode, does not need to gather a large amount of flag datas in advance, saves time and manpower, and improve recognition capability and the level of intelligence of robot, have very strong practicality.
Below in conjunction with accompanying drawing illustrated embodiment, a kind of child's knowledge figure based on man-machine interaction of the present invention and data extending method are described in further detail:
Accompanying drawing explanation
Fig. 1 is the product process figure of category-A data in a kind of child's knowledge figure based on man-machine interaction of the present invention and data extending method;
Fig. 2 is the frame line chart obtaining category-B data in a kind of child's knowledge figure based on man-machine interaction of the present invention and data extending method.
Fig. 3 is the frame line chart that in a kind of child's knowledge figure based on man-machine interaction of the present invention and data extending method, man-machine interaction participant is loaded into category-A data or category-B data at random by client unit;
Fig. 4 is the process flow diagram that in a kind of child's knowledge figure based on man-machine interaction of the present invention and data extending method, man-machine interaction participant carries out man-machine interaction by client unit;
Fig. 5 is be the frame line chart of category-A data by category-B data transformations in a kind of child's knowledge figure based on man-machine interaction of the present invention and data extending method.
Embodiment
First it should be noted that, in the present patent application file, described child and man-machine interaction participant should make same concept and understand, and all refer to the personnel being participated in man-machine interaction by client unit, namely described robot refers to client unit.A kind of child based on man-machine interaction of the present invention knows figure and data extending method, man-machine interactive platform comprises server unit and multiple client unit, server unit is communicated by wired or wireless mode with client unit, and child's knowledge figure and data extending method comprise the following steps:
One, by initialization data stored in the database of server unit, initialization data comprises subject image and corresponding object names; Initialization data is divided into category-A data also for each subject image of category-A data distributes unique number, described numbering is for identifying this subject image;
Two, by client unit shot object image or from Network Capture subject image, will take or upload onto the server the database of unit from the subject image of Network Capture; Be divided into category-B data also for each subject image of category-B data distributes unique number by shooting or from the subject image of Network Capture, described numbering is for identifying this subject image;
Three, man-machine interaction participant is loaded into the subject image of category-A data in server unit database or category-B data at random by client unit, and answers corresponding object names according to the subject image be loaded into;
Four, for the subject image number of interactive process loading category-A data, answer correctness according to man-machine interaction participant calculates gained achievement score, and records the object names of man-machine interaction participant for each subject image answer of category-B data;
Five, achievement score step 4 obtained and man-machine interaction participant to upload onto the server unit in the lump for the object names that each subject image of category-B data is answered, as the man-machine interaction result that this interactive process returns;
Six, according to step one to the mode of step 5, upload onto the server all results participating in interactive process by client unit unit;
Seven, for the jobbie image of category-B data, according to repeatedly or the result of multiple interactive process calculate the weights of each interactive process gained achievement score respectively, and determine the final object names of this subject image according to the size of weights;
The final object names of certain the category-B data subject image eight, obtained according to step 7, marks the subject image of these category-B data, is category-A data by it from category-B data transformations.
In above-mentioned steps four, the described subject image number for interactive process loading category-A data, the answer correctness according to man-machine interaction participant calculates gained achievement score, specifically comprises the following steps:
(1) be loaded into the subject image sum of category-A data by accumulation calculating this interactive process, and be recorded as m;
(2) according to being loaded into the numbering of often opening the subject image of category-A data, the object names answer man-machine interaction participant and the object names of this subject image corresponding stored in a database compare, if consistent, be judged as answering correctly, if inconsistent, be judged as erroneous answers, statistics answers correct number of times, and is recorded as m
1
(3) with answering the subject image total number of correct number of times divided by being loaded into category-A data, using gained divisor as achievement score, and s=m is designated as
1/ m.
In above-mentioned steps seven, described basis repeatedly or the result of multiple interactive process calculate the weights of each interactive process gained achievement score respectively, and determine the final object names of this subject image according to the size of weights, specifically comprise the following steps:
(1) for a certain subject image of category-B data, the number of times returning man-machine interaction result is set to n, achievement score n interactive process returned is set to set s
i={ s
1, s
2, s
3..., s
n; Man-machine interaction participant n interactive process returned is set to set d for the object names that this subject image is answered
i={ d
1, d
2, d
3..., d
n;
(2) object names answered for this subject image according to the whether identical man-machine interaction participant returned n interactive process of content is distinguished, and obtains the different object names of k kind, be designated as and gather d according to the identical content mode only retained once
j={ d
1, d
2, d
3..., d
k.
(3) according to set d
j={ d
1, d
2, d
3..., d
kin each element, with d
j=d
ifor the achievement score that each element of condition accumulation calculating is corresponding, as the mark weight w that each element is corresponding
i,
w
i＝∑(s
i|d
j＝d
i)
(4) corresponding to each element mark weight w
ibe normalized, as the normalization weights that each element is corresponding
(5) for the result that step (four) obtains, normalization weights corresponding for each element are sorted from big to small and add up successively, until added the threshold value t that accumulative normalization weights sum is greater than setting, wherein 0 < t < 1;
(6) judge according to the number adding accumulative normalization weights in step (five),
If add the number j=1 of accumulative normalization weights:, direct by element d corresponding for maximum normalization weights
jas the final object names to this subject image;
If add the number j=2 of accumulative normalization weights, 3: the method for then being analysed and compared by the meaning of a word judges corresponding several element d
jwhether be synonym, if synonym or the word for comprising justice, then simultaneously by these elements d
jas the final object names of this subject image;
If add the number j=2 of accumulative normalization weights, several element d corresponding when 3
jit is not synonym, or when adding the number j>3 of accumulative normalization weights, the condition that then represents does not meet cannot determine final object names, and by increasing the number of times returning man-machine interaction result, continue to perform step () to step (five), until determine the final object names of this subject image.
As embodiment, the value of described threshold value t can be set to 0.7.It should be noted that, the value of threshold value t can adjust according to accuracy requirement, and the larger final object names determined of value is more accurate, and practical application shows, t value 0.7 can realize the object of the invention.In practical application, allow the corresponding one or more object names of category-A data subject image, and allow an object names multiple subject image corresponding; The subject image of category-A data and category-B data all adopts the image of fixed size, and makes object be positioned at image centre position, and stores with jpg, bmp form; The object names of category-A data can adopt one or more Chinese-character words to represent.
For helping skilled in the art to understand the present invention, in the mode of specific embodiment, a kind of child's knowledge figure based on man-machine interaction of the present invention and data extending method are further described below.
Method of the present invention, for child's early education, is carried out with the man-machine interaction mode of looking-and-understanding.Man-machine interactive platform comprises server unit and multiple client unit; Child's knowledge figure and data extending method are made up of data initialization, man-machine interaction, data processing three processes, and detailed process is:
Before use category-A data are stored in the database of server unit, before use or in using category-B data are also stored in the database of server unit.
Category-A data comprise subject image+object names, namely have marker samples;
Category-B data only have subject image, i.e. unmarked sample;
Wherein, subject image employing background is simple, the image of fixed size, and allows object be positioned at the centre position of image, stores with image formats such as jpg, Bmp; Object names adopts one or more Chinese-character words to represent.
For category-A data, classifying and make each object names descriptor multiple different subject image corresponding by object names, to each subject image, is all that it distributes unique number and makes its one or more different object names corresponding.As shown in Figure 1, the generation standard of category-A data can be carried out as follows:
I. input is started;
Ii. take not containing the background image I of object
0, also can directly adopt simple background;
Iii. I is taken
0the image I of object is comprised under background
1;
Iv. subject image I=I
1-I
0;
V. speech recognition or mode word input object names, form one or several word as object names descriptor.
For category-B data, as shown in Figure 2, directly can obtain image from internet, also can take image by the robot as client in daily routines, and be uploaded onto the server, the subject image being each category-B data by server distributes unique number.
As shown in Figure 3 and Figure 4, in interactive process, the category-A data that the robot as client unit reads and category-B data are all from the database of server unit, and detailed process is as follows:
(1) start interactive process, namely know figure process, robot is loaded into data at random and shows image, and put question to " what this is ", loaded data both can be category-A data, also can be category-B data;
(2) man-machine interaction participant answers object names by voice mode, and speech recognition is become word by speech recognition by robot, as answer;
(3) if be loaded into be category-A data:, answer and object names are compared, provide correct/error adjudicate;
(4) if what be loaded into is category-B data:, will directly by answer record;
(5) at an interactive process, be m by the subject image number scale of loading category-A data, identify that correct number is designated as m
1, then the achievement score of this interactive process is: s=100 × m
1/ m.
After one time interactive process terminates, by the identification gained achievement score s for category-A data subject image, and each opens the corresponding answer upload of subject image to server unit to category-B data.
As shown in Figure 5, the database of server unit comprises the data that all client unit (robot) use, multiple stage robot is connected to same server unit by mobile Internet, by the achievement score of man-machine interaction each time and to the answer upload of category-B data to server unit, allow server unit according to repeatedly or the result of multiple interactive process calculate the weights of each interactive process gained achievement score respectively, and the size of foundation weights determines the accurate answer of this subject image, as final object names, concrete grammar is as follows:
For a certain subject image of category-B data, the number of times returning man-machine interaction result is set to n, mark n interactive process returned is set to set s
i, s
i={ s
1, s
2, s
3..., s
n; Man-machine interaction participant n interactive process returned is set to set d for the object names that this subject image is answered
i, d
i={ d
1, d
2, d
3..., d
n.
The object names answered for this subject image according to the whether identical man-machine interaction participant returned n interactive process of content is distinguished, and obtains the different object names of k kind, be designated as and gather d according to the identical content mode only retained once
j={ d
1, d
2, d
3..., d
k.
According to set d
j={ d
1, d
2, d
3..., d
kin each element, with d
j=d
ifor the achievement score that each element of condition accumulation calculating is corresponding, as the mark weights that each element is corresponding, be designated as w
i.
w
i＝∑(s
i|d
j＝d
i)
The mark weight w corresponding to each element
ibe normalized, as the normalization weights that each element is corresponding, be designated as
Normalization weights corresponding for each element are sorted from big to small and add up successively, until added the threshold value t that accumulative normalization weights sum is greater than setting, wherein 0 < t < 1; T value is larger, and the final object names determined is more accurate, and practical application shows, t value 0.7 can realize the object of the invention.
Judge, if add the number j=1 of accumulative normalization weights according to the number adding accumulative normalization weights:, direct by element d corresponding for maximum normalization weights
jas the final object names to this subject image;
If add the number j=2 of accumulative normalization weights, 3: the method for then being analysed and compared by the meaning of a word judges corresponding several element d
jwhether be synonym, if synonym or the word (as automobile, car) for comprising justice, then simultaneously by these elements d
jas the final object names of this subject image;
If add the number j=2 of accumulative normalization weights, several element d corresponding when 3
jbe not synonym (as automobile, aircraft), or when adding the number j>3 of accumulative normalization weights, the condition that then represents does not meet cannot determine final object names, and by increasing the number of times returning man-machine interaction result, continue according to said method until determine the final object names of this subject image.
According to the final object names determined, the subject image of these category-B data is marked, with by its from category-B data transformations for category-A data.
Above embodiment is only the description carried out the preferred embodiment of the present invention; the restriction not request protection domain of the present invention carried out; under the prerequisite not departing from design concept of the present invention and spirit; the various forms of distortion that this area engineering technical personnel make according to technical scheme of the present invention, all should fall in the determined protection domain of claims of the present invention.
Claims (7)
1. the child based on man-machine interaction knows figure and data extending method, described man-machine interactive platform comprises server unit and multiple client unit, server unit is communicated by wired or wireless mode with client unit, and its feature exists: described child's knowledge figure and data extending method comprise the following steps:
One, by initialization data stored in the database of server unit, initialization data comprises subject image and corresponding object names; Initialization data is divided into category-A data also for each subject image of category-A data distributes unique number, described numbering is for identifying this subject image;
Two, by client unit shot object image or from Network Capture subject image, will take or upload onto the server the database of unit from the subject image of Network Capture; Be divided into category-B data also for each subject image of category-B data distributes unique number by shooting or from the subject image of Network Capture, described numbering is for identifying this subject image;
Three, man-machine interaction participant is loaded into the subject image of category-A data in server unit database or category-B data at random by client unit, and answers corresponding object names according to the subject image be loaded into;
Four, for the subject image number of interactive process loading category-A data, answer correctness according to man-machine interaction participant calculates gained achievement score, and records the object names of man-machine interaction participant for each subject image answer of category-B data;
Five, achievement score step 4 obtained and man-machine interaction participant to upload onto the server unit in the lump for the object names that each subject image of category-B data is answered, as the man-machine interaction result that this interactive process returns;
Six, according to step one to the mode of step 5, upload onto the server all results participating in interactive process by client unit unit;
Seven, for the jobbie image of category-B data, according to repeatedly or the result of multiple interactive process calculate the weights of each interactive process gained achievement score respectively, and determine the final object names of this subject image according to the size of weights;
The final object names of certain the category-B data subject image eight, obtained according to step 7, marks the subject image of these category-B data, is category-A data by it from category-B data transformations.
2. know figure and data extending method according to a kind of child based on man-machine interaction according to claim 1, it is characterized in that: in above-mentioned steps four, the described subject image number for interactive process loading category-A data, answer correctness according to man-machine interaction participant calculates gained achievement score, specifically comprises the following steps:
(1) be loaded into the subject image sum of category-A data by accumulation calculating this interactive process, and be recorded as m;
(2) according to being loaded into the numbering of often opening the subject image of category-A data, the object names answer man-machine interaction participant and the object names of this subject image corresponding stored in a database compare, if consistent, be judged as answering correctly, if inconsistent, be judged as erroneous answers, statistics answers correct number of times, and is recorded as m
1
(3) with answering the subject image total number of correct number of times divided by being loaded into category-A data, using gained divisor as achievement score, and s=m is designated as
1/ m.
3. know figure and data extending method according to a kind of child based on man-machine interaction according to claim 2, it is characterized in that: in above-mentioned steps seven, described basis repeatedly or the result of multiple interactive process calculate the weights of each interactive process gained achievement score respectively, and the size of foundation weights determines the final object names of this subject image, specifically comprises the following steps:
(1) for a certain subject image of category-B data, the number of times returning man-machine interaction result is set to N, achievement score N interactive process returned is set to set s
i={ s
1, s
2, s
3..., s
n; Man-machine interaction participant N interactive process returned is set to set d for the object names that this subject image is answered
i={ d
1, d
2, d
3..., d
n;
(2) object names answered for this subject image according to the whether identical man-machine interaction participant returned N interactive process of content is distinguished, and obtains the different object names of K kind, be designated as and gather d according to the identical content mode only retained once
j={ d
1, d
2, d
3..., d
k.
(3) according to set d
j={ d
1, d
2, d
3..., d
kin each element, with d
j=d
ifor the achievement score that each element of condition accumulation calculating is corresponding, as the mark weight w that each element is corresponding
i,
w
i＝∑(s
i|d
j＝d
i)
(4) corresponding to each element mark weight w
ibe normalized, as the normalization weights that each element is corresponding
(5) for the result that step (four) obtains, normalization weights corresponding for each element are sorted from big to small and add up successively, until added the threshold value t that accumulative normalization weights sum is greater than setting, wherein 0 < t < 1;
(6) judge according to the number adding accumulative normalization weights in step (five),
If add the number j=1 of accumulative normalization weights:, direct by element d corresponding for maximum normalization weights
jas the final object names to this subject image;
If add the number j=2 of accumulative normalization weights, 3: the method for then being analysed and compared by the meaning of a word judges corresponding several element d
jwhether be synonym, if synonym or the word for comprising justice, then simultaneously by these elements d
jas the final object names of this subject image;
If add the number j=2 of accumulative normalization weights, several element d corresponding when 3
jit is not synonym, or when adding the number j>3 of accumulative normalization weights, the condition that then represents does not meet cannot determine final object names, and by increasing the number of times returning man-machine interaction result, continue to perform step () to step (five), until determine the final object names of this subject image.
4. know figure and data extending method according to a kind of child based on man-machine interaction according to claim 3, it is characterized in that: the value of described threshold value t is 0.7.
5. know figure and data extending method according to a kind of child based on man-machine interaction according to claim 1, it is characterized in that: a corresponding one or more object names of subject image in described category-A data, and object names multiple subject image corresponding.
6. know figure and data extending method according to a kind of child based on man-machine interaction according to claim 1, it is characterized in that: the subject image of described category-A data and category-B data all adopts the image of fixed size, and make object be positioned at image centre position, and store with jpg, bmp form.
7. know figure and data extending method according to a kind of child based on man-machine interaction according to claim 1, it is characterized in that: the object names of described category-A data adopts one or more Chinese-character words to represent.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510862736.4A CN105469116B (en) | 2015-12-01 | 2015-12-01 | A kind of child's knowledge figure and data extending method based on human-computer interaction |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510862736.4A CN105469116B (en) | 2015-12-01 | 2015-12-01 | A kind of child's knowledge figure and data extending method based on human-computer interaction |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105469116A true CN105469116A (en) | 2016-04-06 |
| CN105469116B CN105469116B (en) | 2019-01-04 |
Family
ID=55606787
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510862736.4A Active CN105469116B (en) | 2015-12-01 | 2015-12-01 | A kind of child's knowledge figure and data extending method based on human-computer interaction |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105469116B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106873893A (en) * | 2017-02-13 | 2017-06-20 | 北京光年无限科技有限公司 | For the multi-modal exchange method and device of intelligent robot |
| CN107589963A (en) * | 2017-09-26 | 2018-01-16 | 维沃移动通信有限公司 | A kind of image processing method, mobile terminal and computer-readable recording medium |
| CN112464894A (en) * | 2020-12-14 | 2021-03-09 | 深圳市优必选科技股份有限公司 | Interaction method and device and computer equipment |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101093572A (en) * | 2007-07-20 | 2007-12-26 | 中国建设银行股份有限公司 | A preposition system and a centralized data processing system |
| US20100241658A1 (en) * | 2005-04-08 | 2010-09-23 | Rathurs Spencer A | System and method for accessing electronic data via an image search engine |
| CN102159134A (en) * | 2008-07-08 | 2011-08-17 | 普罗秋斯生物医学公司 | Ingestible event marker data framework |
| CN103035135A (en) * | 2012-11-27 | 2013-04-10 | 北京航空航天大学 | Children cognitive system based on augment reality technology and cognitive method |
| CN104157168A (en) * | 2014-08-21 | 2014-11-19 | 鲜于庆 | Intelligent database application system applied to artificial intelligent robot and interaction method |
-
2015
- 2015-12-01 CN CN201510862736.4A patent/CN105469116B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100241658A1 (en) * | 2005-04-08 | 2010-09-23 | Rathurs Spencer A | System and method for accessing electronic data via an image search engine |
| CN101093572A (en) * | 2007-07-20 | 2007-12-26 | 中国建设银行股份有限公司 | A preposition system and a centralized data processing system |
| CN102159134A (en) * | 2008-07-08 | 2011-08-17 | 普罗秋斯生物医学公司 | Ingestible event marker data framework |
| CN103035135A (en) * | 2012-11-27 | 2013-04-10 | 北京航空航天大学 | Children cognitive system based on augment reality technology and cognitive method |
| CN104157168A (en) * | 2014-08-21 | 2014-11-19 | 鲜于庆 | Intelligent database application system applied to artificial intelligent robot and interaction method |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106873893A (en) * | 2017-02-13 | 2017-06-20 | 北京光年无限科技有限公司 | For the multi-modal exchange method and device of intelligent robot |
| CN106873893B (en) * | 2017-02-13 | 2021-01-22 | 北京光年无限科技有限公司 | Multi-modal interaction method and device for intelligent robot |
| CN107589963A (en) * | 2017-09-26 | 2018-01-16 | 维沃移动通信有限公司 | A kind of image processing method, mobile terminal and computer-readable recording medium |
| CN107589963B (en) * | 2017-09-26 | 2019-05-17 | 维沃移动通信有限公司 | A kind of image processing method, mobile terminal and computer readable storage medium |
| CN112464894A (en) * | 2020-12-14 | 2021-03-09 | 深圳市优必选科技股份有限公司 | Interaction method and device and computer equipment |
| CN112464894B (en) * | 2020-12-14 | 2023-09-01 | 深圳市优必选科技股份有限公司 | Interaction method and device and computer equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105469116B (en) | 2019-01-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2020238054A1 (en) | Method and apparatus for positioning chart in pdf document, and computer device | |
| CN109800320B (en) | Image processing method, device and computer readable storage medium | |
| CN112926405A (en) | Method, system, equipment and storage medium for detecting wearing of safety helmet | |
| CN110059157A (en) | A kind of picture and text cross-module state search method, system, device and storage medium | |
| CN105654066A (en) | Vehicle identification method and device | |
| CN107679475A (en) | Shops's monitoring evaluation method, device and storage medium | |
| CN105224920A (en) | A kind ofly will proceed to the learning test diagnostic system on line under line | |
| CN109284700B (en) | Method, storage medium, device and system for detecting multiple faces in image | |
| CN109800309A (en) | Classroom Discourse genre classification methods and device | |
| CN111144079A (en) | Method and device for intelligently acquiring learning resources, printer and storage medium | |
| CN105469116A (en) | Picture recognition and data extension method for infants based on man-machine interaction | |
| CN111444339A (en) | Text question difficulty labeling method and device and computer readable storage medium | |
| CN114639152A (en) | Multi-modal voice interaction method, device, equipment and medium based on face recognition | |
| CN112995690B (en) | Live content category identification method, device, electronic equipment and readable storage medium | |
| CN111191027B (en) | Generalized zero sample identification method based on Gaussian mixture distribution (VAE) | |
| CN112991410A (en) | Text image registration method, electronic equipment and storage medium thereof | |
| CN116228361A (en) | Course recommendation method, device, equipment and storage medium based on feature matching | |
| CN112131354B (en) | Answer screening method and device, terminal equipment and computer readable storage medium | |
| CN111339939B (en) | Attendance checking method and device based on image recognition | |
| CN114722280A (en) | User portrait based course recommendation method, device, equipment and storage medium | |
| CN114520059A (en) | Traditional Chinese medicine diagnostics data platform based on big data | |
| CN113822589A (en) | Intelligent interviewing method, device, equipment and storage medium | |
| CN111464743A (en) | Photographic composition matching method and system | |
| CN117152846B (en) | Student behavior recognition method, device and system and computer readable storage medium | |
| CN110069783A (en) | A kind of answer content evaluating method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |