CN104834918A - Human behavior recognition method based on Gaussian process classifier - Google Patents
Human behavior recognition method based on Gaussian process classifier Download PDFInfo
- Publication number
- CN104834918A CN104834918A CN201510259853.1A CN201510259853A CN104834918A CN 104834918 A CN104834918 A CN 104834918A CN 201510259853 A CN201510259853 A CN 201510259853A CN 104834918 A CN104834918 A CN 104834918A
- Authority
- CN
- China
- Prior art keywords
- gaussian process
- sorter
- test data
- human body
- human
- 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
- 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
- G06V40/20—Movements or behaviour, e.g. gesture recognition
- G06V40/23—Recognition of whole body movements, e.g. for sport training
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- 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
- G06V40/103—Static body considered as a whole, e.g. static pedestrian or occupant recognition
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Multimedia (AREA)
- Human Computer Interaction (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Psychiatry (AREA)
- Social Psychology (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention provides a human behavior recognition method based on a Gaussian process classifier, which comprises the steps of 1) training a plurality of binary classifier models, wherein each model corresponds to the probability for judging belonging to a certain class of behaviors; 2) acquiring test data of human behaviors, and respectively calculating the probabilities of human behavior classes to which the test data belongs through the Gaussian process binary classifier models; and 3) comparing probability values of the human behavior classes, and determining the human behavior class corresponding to the maximum value to be the human behavior class to which the test data belongs. Compared with a traditional classifier, the Gaussian process classifier can predict the probability of belonging to a certain class of behaviors, various types of weight selectivity can be provided when the class to which the test data belongs is determined. Compared with a classification technique support vector machine which is popular and mature, the Gaussian process classifier has better classification accuracy when the dimensionality of input data is high, and the used training time is obviously shortened when the amount of training data is great.
Description
Technical field
The invention belongs to Human Body Model and identify field, particularly relate to a kind of Human bodys' response method based on Gaussian process sorter.
Background technology
Human bodys' response refers to type of action to people, behavior pattern is analyzed and identifies, generally by adopting a kind of suitable mode to express the relevant information extracted from the various types of data sequence of physical world, and explain these information, to realize the behavior learning and identify people.Wearable Activity recognition is mainly through dressing the mode of motion of various sensor Direct Recognition user at health different parts.Due to advantages such as structure are simple, data processing is easy, instrument cost is cheap, Wearable Activity recognition has become important technological means and study hotspot in the fields such as medical science long distance monitoring, health care, emergency management and rescue, military and national defense.
The processes such as the Human bodys' response system based on Wearable sensor generally can be divided into data acquisition, window length is chosen, feature extraction, feature selecting, dimension-reduction treatment, classification, result verification, wherein choosing of sorter is one of gordian technique in whole system, has a direct impact for classification results.At present, conventional sorting algorithm has support vector machine, artificial neural network, Hidden Markov Model (HMM), Bayes classifier etc.But, they respectively have drawback, as Bayes classifier needs a large amount of sample datas, artificial neural network is except needing a large amount of sample datas, its network structure is also difficult to determine, assumed condition in Hidden Markov Model (HMM) is comparatively strict, and practicality is poor, and support vector machine exists regularization coefficient, kernel functional parameter determines the inherent limitations such as difficulty.In addition, traditional sorting algorithm is all difficult to obtain good classifying quality when the dimension inputting data is higher.
Gaussian process sorter is a kind of new statistical learning algorithm based on Gaussian process and bayesian theory, is the hot fields of current machine learning in the world research.Because Gaussian process sorter has the following advantages: hyper parameter is optimized voluntarily in the process of model learning; With the form output category result of probability, be easy to the uncertainty of problem analysis; Effective solution small sample, higher-dimension, the complicated classification problem etc. such as non-linear, in recent years, Gaussian process sorter achieves good achievement in research in the Human bodys' response etc. of recognition of face, transformer fault diagnosis, slope stability learning and memory and view-based access control model.But Gaussian process sorter is still blank so far being applied in the Human bodys' response based on Wearable sensor.
Given this, the present invention proposes Gaussian process sorter and be applied to Human bodys' response based on Wearable sensor, by comparing with comparatively popular support vector machine, Gaussian process sorter achieves obviously classifying quality preferably under higher-dimension input and Small Sample Size.
Summary of the invention
The shortcoming of prior art in view of the above, the object of the present invention is to provide a kind of Human bodys' response method based on Gaussian process sorter, for solving the dissatisfactory problem of human body behaviour classification recognition effect in prior art.
For achieving the above object and other relevant objects, the invention provides a kind of Human bodys' response method based on Gaussian process sorter, comprise step:
1) train several Gaussian process two sorter models, each model corresponds to the probability judging to belong to certain anthropoid behavior;
2) obtain the test data of human body behavior, test data is imported Gaussian process two sorter model that each has trained, calculate the probability of each human body behavior classification belonging to this test data respectively;
3) compare the value of the probability of each human body behavior classification, namely the human body behavior classification corresponding to maximal value is judged to be the human body behavior classification belonging to this test data.
As a kind of preferred version of the Human bodys' response method based on Gaussian process sorter of the present invention, step 1) in, for distinguishing data set corresponding to every anthropoid behavior, the training dataset of input is expressed as: the input data X belonging to the anthropoid behavior of i
i=[x
i1..., x
in]
t, its class is designated as Y
i=[y
i1..., y
in]
t, wherein x
ijrepresent a jth input vector, y
ij∈+1 ,-1}, (i=A, B, C, D J=1 ..., n).
As a kind of preferred version of the Human bodys' response method based on Gaussian process sorter of the present invention, step 1) in, when training Gaussian process two sorter model A, input data are X=[x
a1..., x
an]
t, corresponding class is designated as Y=[y
a1..., y
an]
t, wherein y
aj=+1, y
ij=-1 (i=B, C, D J=1 ..., n), the corresponding class mark namely belonging to the input data of category-A human body behavior is labeled as+1, all the other classifications be labeled as-1, X, Y are sent into Gaussian process two sorter model, train Gaussian process two sorter model A.
As a kind of preferred version of the Human bodys' response method based on Gaussian process sorter of the present invention, step 1) in, train the Gaussian process two sorter model i corresponding to each anthropoid behavior according to the method training Gaussian process two sorter model A, wherein, i=B, C, D ...
As a kind of preferred version of the Human bodys' response method based on Gaussian process sorter of the present invention, step 2) in, the test data of human body behavior is sent into respectively the Gaussian process two sorter model i trained, calculate the probable value P that test data belongs to the anthropoid behavior of i
i, wherein, i=A, B, C, D ...
As a kind of preferred version of the Human bodys' response method based on Gaussian process sorter of the present invention, step 3) in, compare the probable value P of each human body behavior classification
i, namely the human body behavior classification corresponding to maximal value is judged to be the human body behavior classification belonging to this test data, wherein, and i=A, B, C, D ...
As a kind of preferred version of the Human bodys' response method based on Gaussian process sorter of the present invention, described Gaussian process two sorter model comprises Gaussian process two sorter model corresponding respectively to away, run, stand and lie.
As mentioned above, the invention provides a kind of Human bodys' response method based on Gaussian process sorter, comprise step: 1) train several Gaussian process two sorter models, each model corresponds to the probability judging to belong to certain anthropoid behavior; 2) obtain the test data of human body behavior, test data is imported Gaussian process two sorter model that each has trained, calculate the probability of each human body behavior classification belonging to this test data respectively; 3) compare the value of the probability of each human body behavior classification, namely the human body behavior classification corresponding to maximal value is judged to be the human body behavior classification belonging to this test data.The present invention has following beneficial effect:
1) compared with traditional classifier, Gaussian process sorter can dope the probability size belonging to a certain class behavior, and this, when adjudicating the classification belonging to test data, can have multiple weight selectivity.
2) compared with the comparatively popular and sorting technique support vector machine of maturation, Gaussian process sorter, when the dimension inputting data is higher, has better nicety of grading.
3) compared with support vector machine, when training data is more, Gaussian process sorter training time used is obvious much short.
Accompanying drawing explanation
Fig. 1 is shown as the training schematic flow sheet based on Gaussian process two sorter model in the Human bodys' response method of Gaussian process sorter of the present invention.
Fig. 2 learning decision schematic flow sheet based on Gaussian process two sorter model in the Human bodys' response method of Gaussian process sorter of the present invention.
Fig. 3 is shown as Gaussian process sorter of the present invention and the accuracy of identification correlation curve figure of support vector machine when the dimension inputting data is more.
Fig. 4 is shown as the comparative graph of Gaussian process sorter of the present invention and support vector machine training time when training sample is more.
Embodiment
Below by way of specific instantiation, embodiments of the present invention are described, those skilled in the art the content disclosed by this instructions can understand other advantages of the present invention and effect easily.The present invention can also be implemented or be applied by embodiments different in addition, and the every details in this instructions also can based on different viewpoints and application, carries out various modification or change not deviating under spirit of the present invention.
Refer to Fig. 1 ~ Fig. 4.It should be noted that, the diagram provided in the present embodiment only illustrates basic conception of the present invention in a schematic way, then only the assembly relevant with the present invention is shown in graphic but not component count, shape and size when implementing according to reality is drawn, it is actual when implementing, and the kenel of each assembly, quantity and ratio can be a kind of change arbitrarily, and its assembly layout kenel also may be more complicated.
As shown in Figure 1 to 4, the present embodiment provides a kind of Human bodys' response method based on Gaussian process sorter, comprises step:
As shown in Figure 1, first carry out step 1), train several Gaussian process two sorter models, each model corresponds to the probability judging to belong to certain anthropoid behavior.
Exemplarily, step 1) in, for distinguishing data set corresponding to every anthropoid behavior, the training dataset of input is expressed as: the input data X belonging to the anthropoid behavior of i
i=[x
i1..., x
in]
t, its class is designated as Y
i=[y
i1..., y
in]
t, wherein x
ijrepresent a jth input vector, y
ij∈+1 ,-1}, (i=A, B, C, D J=1 ..., n).
Exemplarily, step 1) in, when training Gaussian process two sorter model A, input data are X=[x
a1..., x
an]
t, corresponding class is designated as Y=[y
a1..., y
an]
t, wherein y
aj=+1, y
ij=-1 (i=B, C, D J=1 ..., n), the corresponding class mark namely belonging to the input data of category-A human body behavior is labeled as+1, all the other classifications be labeled as-1, X, Y are sent into Gaussian process two sorter model, train Gaussian process two sorter model A.
Exemplarily, step 1) in, train the Gaussian process two sorter model i corresponding to each anthropoid behavior according to the method training Gaussian process two sorter model A, wherein, i=B, C, D ...
Particularly, the gordian technique realizing the object of the invention introduces in Human bodys' response algorithm by Gaussian process sorter, based on Gaussian process two classifier algorithm, adopts one-to-many strategy, realize multiclass Activity recognition.In the present embodiment, to distinguish away, run, stand, four class behaviors of lying, its implementation procedure comprises the steps (wherein Gaussian process two disaggregated model is referred to as two disaggregated models):
Step 1-1), for distinguishing data set corresponding to every anthropoid behavior, the training dataset of input can be expressed as: the input data X belonging to " walking "
a=[x
a1..., x
an]
t, its class is designated as Y
a=[y
a1..., y
an]
t; Belong to the input data X of " race "
b=[x
b1..., x
bn]
t, its class is designated as Y
b=[y
b1..., y
bn]
t; Belong to the input data X of " station "
c=[x
c1..., x
cn]
t, its class is designated as Y
c=[y
c1..., y
cn]
t; Belong to the input data X of " lying "
d=[x
d1..., x
dn]
t, its class is designated as Y
d=[y
d1..., y
dn]
t.Wherein x
ijrepresentation dimension is the input vector of d, y
ij∈ {+1 ,-1} (i=A, B, C, D; J=1 ..., n).
Step 1-2), when training two disaggregated model A, namely train calculating to belong to two disaggregated models of the probability of " walking ".Input data are X=[X
a, X
b, X
c, X
d]
t, corresponding class is designated as Y=[Y
a, Y
b, Y
c, Y
d]
t, wherein y
aj=+1, y
ij=-1 (i=B, C, D; J=1 ..., n).As shown in Figure 1, the corresponding class mark belonging to the input data of " walking " is labeled as+1, and remaining is labeled as-1.X, Y are sent into Gaussian process two sorter model, trains two disaggregated model A.
Step 1-3), when training two disaggregated model B, namely train calculating to belong to two disaggregated models of the probability of " race ".Input data are X=[X
b, X
a, X
c, X
d]
t, corresponding class is designated as Y=[Y
b, Y
a, Y
c, Y
d]
t, wherein y
bj=+1, y
ij=-1 (i=A, C, D; J=1 ..., n).As shown in Figure 1, the corresponding class mark belonging to the input data of " race " is labeled as+1, and remaining is labeled as-1.X, Y are sent into Gaussian process two sorter model, trains two disaggregated model B.
Step 1-4), when training two disaggregated model C, namely train calculating to belong to two disaggregated models of the probability of " station ".Input data are X=[X
c, X
a, X
b, X
d]
t, corresponding class is designated as Y=[Y
c, Y
a, Y
b, Y
d]
t, wherein y
cj=+1, y
ij=-1 (i=A, B, D; J=1 ..., n).As shown in Figure 1, the corresponding class mark belonging to the input data of " station " is labeled as+1, and remaining is labeled as-1.X, Y are sent into Gaussian process two sorter model, trains two disaggregated model C.
Step 1-5), when training two disaggregated model D, namely train calculating to belong to two disaggregated models of the probability of " lying ".Input data are X=[X
d, X
a, X
b, X
c]
t, corresponding class is designated as Y=[Y
d, Y
a, Y
b, Y
c]
t, wherein y
dj=+1, y
ij=-1 (i=A, B, C; J=1 ..., n).As shown in Figure 1, the corresponding class mark belonging to the input data of " lying " is labeled as+1, and remaining is labeled as-1.X, Y are sent into Gaussian process two sorter model, trains two disaggregated model D.
As shown in Figure 2, then carry out step 2), obtain the test data of human body behavior, test data is imported Gaussian process two sorter model that each has trained, calculate the probability of each human body behavior classification belonging to this test data respectively.
Exemplarily, step 2) in, the test data of human body behavior is sent into respectively the Gaussian process two sorter model i trained, calculate the probable value P that test data belongs to the anthropoid behavior of i
i, wherein, i=A, B, C, D ...
As shown in Figure 2, particularly, to test data x, calculate the classification belonging to it, in the present embodiment, comprise the following steps:
Step 2-1), first x is fed in step 1-2) in the two disaggregated model A that trained, calculate the probable value P that test data x belongs to " walking "
a.
Step 2-2), then x is fed in step 1-3) in the two disaggregated model B that trained, calculate the probable value P that test data x belongs to " race "
b.
Step 2-3), then x is fed in step 1-4) in the two disaggregated model C that trained, calculate the probable value P that test data x belongs to " station "
c.
Step 2-4), finally x is fed in step 1-5) in the two disaggregated model D that trained, calculate the probable value P that test data x belongs to " lying "
d.
Finally, as shown in Figure 2, carry out step 3), compare the value of the probability of each human body behavior classification, namely the human body behavior classification corresponding to maximal value is judged to be the human body behavior classification belonging to this test data.
Exemplarily, step 3) in, compare the probable value P of each human body behavior classification
i, namely the human body behavior classification corresponding to maximal value is judged to be the human body behavior classification belonging to this test data, wherein, and i=A, B, C, D ...
Particularly, as shown in Figure 2, P is compared
a, P
b, P
c, P
dvalue, namely the classification corresponding to maximal value is judged to the human body behavior classification belonging to test data x.
Fig. 3 is shown as Gaussian process sorter of the present invention and the accuracy of identification correlation curve figure of support vector machine when the dimension inputting data is more, as can be seen from curve map, compared with the comparatively popular and sorting technique support vector machine of maturation, Gaussian process sorter of the present invention, when the dimension inputting data x is higher, has better nicety of grading.
Fig. 4 is shown as the comparative graph of Gaussian process sorter of the present invention and support vector machine training time when training sample is more, as can be seen from curve map, compared with support vector machine, when training data is more, Gaussian process sorter of the present invention training time used is obvious much short.
As mentioned above, the invention provides a kind of Human bodys' response method based on Gaussian process sorter, comprise step: 1) train several Gaussian process two sorter models, each model corresponds to the probability judging to belong to certain anthropoid behavior; 2) obtain the test data of human body behavior, test data is imported Gaussian process two sorter model that each has trained, calculate the probability of each human body behavior classification belonging to this test data respectively; 3) compare the value of the probability of each human body behavior classification, namely the human body behavior classification corresponding to maximal value is judged to be the human body behavior classification belonging to this test data.The present invention has following beneficial effect:
1) compared with traditional classifier, Gaussian process sorter can dope the probability size belonging to a certain class behavior, and this, when adjudicating the classification belonging to test data, can have multiple weight selectivity.
2) compared with the comparatively popular and sorting technique support vector machine of maturation, Gaussian process sorter, when the dimension inputting data is higher, has better nicety of grading.
3) compared with support vector machine, when training data is more, Gaussian process sorter training time used is obvious much short.
So the present invention effectively overcomes various shortcoming of the prior art and tool high industrial utilization.
Above-described embodiment is illustrative principle of the present invention and effect thereof only, but not for limiting the present invention.Any person skilled in the art scholar all without prejudice under spirit of the present invention and category, can modify above-described embodiment or changes.Therefore, such as have in art usually know the knowledgeable do not depart from complete under disclosed spirit and technological thought all equivalence modify or change, must be contained by claim of the present invention.
Claims (7)
1., based on a Human bodys' response method for Gaussian process sorter, it is characterized in that, comprise step:
1) train several Gaussian process two sorter models, each model corresponds to the probability judging to belong to certain anthropoid behavior;
2) obtain the test data of human body behavior, test data is imported Gaussian process two sorter model that each has trained, calculate the probability of each human body behavior classification belonging to this test data respectively;
3) compare the value of the probability of each human body behavior classification, namely the human body behavior classification corresponding to maximal value is judged to be the human body behavior classification belonging to this test data.
2. the Human bodys' response method based on Gaussian process sorter according to claim 1, it is characterized in that: step 1) in, for distinguishing data set corresponding to every anthropoid behavior, the training dataset of input is expressed as: the input data X belonging to the anthropoid behavior of i
i=[x
i1..., x
in]
t, its class is designated as Y
i=[y
i1..., y
in]
t, wherein x
ijrepresent a jth input vector, y
ij∈+1 ,-1}, (i=A, B, C, D J=1 ..., n).
3. the Human bodys' response method based on Gaussian process sorter according to claim 2, is characterized in that: step 1) in, when training Gaussian process two sorter model A, input data are X=[x
a1..., x
an]
t, corresponding class is designated as Y=[y
a1..., y
an]
t, wherein y
aj=+1, y
ij=-1 (i=B, C, D J=1 ..., n), the corresponding class mark namely belonging to the input data of category-A human body behavior is labeled as+1, all the other classifications be labeled as-1, X, Y are sent into Gaussian process two sorter model, train Gaussian process two sorter model A.
4. the Human bodys' response method based on Gaussian process sorter according to claim 3, it is characterized in that: step 1) in, the Gaussian process two sorter model i corresponding to each anthropoid behavior is trained according to the method training Gaussian process two sorter model A, wherein, i=B, C, D ...
5. the Human bodys' response method based on Gaussian process sorter according to claim 4, it is characterized in that: step 2) in, the test data of human body behavior is sent into respectively the Gaussian process two sorter model i trained, calculate the probable value P that test data belongs to the anthropoid behavior of i
i, wherein, i=A, B, C, D ...
6. the Human bodys' response method based on Gaussian process sorter according to claim 5, is characterized in that: step 3) in, compare the probable value P of each human body behavior classification
i, namely the human body behavior classification corresponding to maximal value is judged to be the human body behavior classification belonging to this test data, wherein, and i=A, B, C, D ...
7. the Human bodys' response method based on Gaussian process sorter according to claim 1 ~ 6 any one, is characterized in that: described Gaussian process two sorter model comprises Gaussian process two sorter model corresponding respectively to away, run, stand and lie.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510259853.1A CN104834918A (en) | 2015-05-20 | 2015-05-20 | Human behavior recognition method based on Gaussian process classifier |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510259853.1A CN104834918A (en) | 2015-05-20 | 2015-05-20 | Human behavior recognition method based on Gaussian process classifier |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104834918A true CN104834918A (en) | 2015-08-12 |
Family
ID=53812796
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510259853.1A Pending CN104834918A (en) | 2015-05-20 | 2015-05-20 | Human behavior recognition method based on Gaussian process classifier |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104834918A (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105426913A (en) * | 2015-11-16 | 2016-03-23 | 南京工程学院 | Method and device for constructing Gaussian process multi-classifier |
| CN106056093A (en) * | 2016-06-13 | 2016-10-26 | 西安电子科技大学 | Human behavior recognition method based on binary-classification evolution characteristic learning |
| CN106932200A (en) * | 2017-04-20 | 2017-07-07 | 北京信息科技大学 | The Gaussian process multiple features intelligent diagnosing method of rolling bearing fault |
| WO2017140222A1 (en) * | 2016-02-19 | 2017-08-24 | 阿里巴巴集团控股有限公司 | Modelling method and device for machine learning model |
| CN108090503A (en) * | 2017-11-28 | 2018-05-29 | 东软集团股份有限公司 | On-line tuning method, apparatus, storage medium and the electronic equipment of multi-categorizer |
| CN108182339A (en) * | 2018-03-20 | 2018-06-19 | 北京工业大学 | A kind of window trend prediction method and system based on Gaussian Profile |
| CN109314840A (en) * | 2016-06-27 | 2019-02-05 | 英特尔公司 | It is arranged using machine learning optimization wearable device |
| CN109583335A (en) * | 2018-11-16 | 2019-04-05 | 中山大学 | A kind of video human Activity recognition method based on Spatial-temporal Information Fusion |
| CN113989857A (en) * | 2021-12-27 | 2022-01-28 | 四川新网银行股份有限公司 | Portrait photo content analysis method and system based on deep learning |
| CN114881110A (en) * | 2022-04-02 | 2022-08-09 | 西安交通大学 | Real-time detection method for total pressure change mode in on-orbit spacecraft cabin |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CA2704107A1 (en) * | 2009-05-13 | 2010-11-13 | The University Of Sydney | A method and system for data analysis and synthesis |
-
2015
- 2015-05-20 CN CN201510259853.1A patent/CN104834918A/en active Pending
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CA2704107A1 (en) * | 2009-05-13 | 2010-11-13 | The University Of Sydney | A method and system for data analysis and synthesis |
Non-Patent Citations (4)
| Title |
|---|
| HANG ZHOU等: "Human action recognition by feature-reduced Gaussian process classification", 《PATTERN RECOGNITION LETTERS》 * |
| XIANGHANG LIU等: "Active learning for human action recognition with Gaussian Processes", 《2011 18TH IEEE INTERNATIONAL CONFERENCE ON IMAGE PROCESSING》 * |
| 薛洋: "基于单个加速度传感器的人体运动模式识别", 《中国博士学位论文全文数据库信息科技辑》 * |
| 魏三喜: "基于高斯过程的多元分类方法", 《信息技术》 * |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105426913A (en) * | 2015-11-16 | 2016-03-23 | 南京工程学院 | Method and device for constructing Gaussian process multi-classifier |
| WO2017140222A1 (en) * | 2016-02-19 | 2017-08-24 | 阿里巴巴集团控股有限公司 | Modelling method and device for machine learning model |
| CN106056093A (en) * | 2016-06-13 | 2016-10-26 | 西安电子科技大学 | Human behavior recognition method based on binary-classification evolution characteristic learning |
| CN106056093B (en) * | 2016-06-13 | 2019-05-21 | 西安电子科技大学 | Human bodys' response method based on two classification Evolution study |
| CN109314840A (en) * | 2016-06-27 | 2019-02-05 | 英特尔公司 | It is arranged using machine learning optimization wearable device |
| CN106932200A (en) * | 2017-04-20 | 2017-07-07 | 北京信息科技大学 | The Gaussian process multiple features intelligent diagnosing method of rolling bearing fault |
| CN108090503A (en) * | 2017-11-28 | 2018-05-29 | 东软集团股份有限公司 | On-line tuning method, apparatus, storage medium and the electronic equipment of multi-categorizer |
| CN108182339A (en) * | 2018-03-20 | 2018-06-19 | 北京工业大学 | A kind of window trend prediction method and system based on Gaussian Profile |
| CN108182339B (en) * | 2018-03-20 | 2021-08-13 | 北京工业大学 | Window state prediction method and system based on Gaussian distribution |
| CN109583335A (en) * | 2018-11-16 | 2019-04-05 | 中山大学 | A kind of video human Activity recognition method based on Spatial-temporal Information Fusion |
| CN113989857A (en) * | 2021-12-27 | 2022-01-28 | 四川新网银行股份有限公司 | Portrait photo content analysis method and system based on deep learning |
| CN114881110A (en) * | 2022-04-02 | 2022-08-09 | 西安交通大学 | Real-time detection method for total pressure change mode in on-orbit spacecraft cabin |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104834918A (en) | Human behavior recognition method based on Gaussian process classifier | |
| CN106709754A (en) | Power user grouping method based on text mining | |
| CN103366367B (en) | Based on the FCM gray-scale image segmentation method of pixel count cluster | |
| CN109886068B (en) | Motion data-based action behavior identification method | |
| CN108564129A (en) | A kind of track data sorting technique based on generation confrontation network | |
| CN105205501B (en) | A kind of weak mark image object detection method of multi classifier combination | |
| CN106445919A (en) | Sentiment classifying method and device | |
| CN106682696A (en) | Multi-example detection network based on refining of online example classifier and training method thereof | |
| CN107578028A (en) | A kind of face identification method, device, equipment and computer-readable recording medium | |
| CN104966105A (en) | Robust machine error retrieving method and system | |
| CN108416364A (en) | Integrated study data classification method is merged in subpackage | |
| CN103500342B (en) | A kind of Human bodys' response method based on accelerometer | |
| CN105930864A (en) | EEG (electroencephalogram) signal feature classification method based on ABC-SVM | |
| CN109086794B (en) | Driving behavior pattern recognition method based on T-LDA topic model | |
| CN105912525A (en) | Sentiment classification method for semi-supervised learning based on theme characteristics | |
| CN103745233B (en) | The hyperspectral image classification method migrated based on spatial information | |
| CN113722474A (en) | Text classification method, device, equipment and storage medium | |
| CN111709468B (en) | Training method and device for directional artificial intelligence and storage medium | |
| CN109255339B (en) | Classification method based on self-adaptive deep forest human gait energy map | |
| Ellis et al. | Physical activity recognition in free-living from body-worn sensors | |
| CN106529377A (en) | Age estimating method, age estimating device and age estimating system based on image | |
| CN108762503A (en) | A kind of man-machine interactive system based on multi-modal data acquisition | |
| CN104978569A (en) | Sparse representation based incremental face recognition method | |
| CN104268587B (en) | False fingerprint detection method based on finger wave conversion and SVM | |
| CN106648149B (en) | A kind of aerial hand-written character recognition method based on accelerometer and gyroscope |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| EXSB | Decision made by sipo to initiate substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20150812 |
|
| RJ01 | Rejection of invention patent application after publication |