CN114612752A - Method and system for intelligently reviewing recognition result of video analysis technology - Google Patents
Method and system for intelligently reviewing recognition result of video analysis technology Download PDFInfo
- Publication number
- CN114612752A CN114612752A CN202210202778.5A CN202210202778A CN114612752A CN 114612752 A CN114612752 A CN 114612752A CN 202210202778 A CN202210202778 A CN 202210202778A CN 114612752 A CN114612752 A CN 114612752A
- Authority
- CN
- China
- Prior art keywords
- result
- rechecking
- recognition
- review
- recognition result
- 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
- 238000004458 analytical method Methods 0.000 title claims abstract description 74
- 238000000034 method Methods 0.000 title claims abstract description 71
- 238000005516 engineering process Methods 0.000 title claims abstract description 65
- 238000012552 review Methods 0.000 claims description 67
- 239000002131 composite material Substances 0.000 claims description 8
- 238000004590 computer program Methods 0.000 claims description 8
- 230000000694 effects Effects 0.000 abstract description 23
- 230000006870 function Effects 0.000 description 7
- 238000012986 modification Methods 0.000 description 6
- 230000004048 modification Effects 0.000 description 6
- 241000282414 Homo sapiens Species 0.000 description 5
- 238000004891 communication Methods 0.000 description 5
- 238000011161 development Methods 0.000 description 5
- 239000003814 drug Substances 0.000 description 5
- 238000012545 processing Methods 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 4
- 238000010801 machine learning Methods 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 238000004422 calculation algorithm Methods 0.000 description 3
- 238000013461 design Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 206010067484 Adverse reaction Diseases 0.000 description 2
- 230000006838 adverse reaction Effects 0.000 description 2
- 238000013529 biological neural network Methods 0.000 description 2
- 210000004556 brain Anatomy 0.000 description 2
- 230000007123 defense Effects 0.000 description 2
- 229940079593 drug Drugs 0.000 description 2
- 238000007726 management method Methods 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 239000000203 mixture Substances 0.000 description 2
- 238000003062 neural network model Methods 0.000 description 2
- 210000002569 neuron Anatomy 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 230000003068 static effect Effects 0.000 description 2
- 238000007418 data mining Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000014509 gene expression Effects 0.000 description 1
- 239000004615 ingredient Substances 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 238000012549 training Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/217—Validation; Performance evaluation; Active pattern learning techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
- G06F18/2155—Generating training patterns; Bootstrap methods, e.g. bagging or boosting characterised by the incorporation of unlabelled data, e.g. multiple instance learning [MIL], semi-supervised techniques using expectation-maximisation [EM] or naïve labelling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/047—Probabilistic or stochastic networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/088—Non-supervised learning, e.g. competitive learning
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Molecular Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Evolutionary Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Probability & Statistics with Applications (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a method and a system for intelligently rechecking a recognition result of a video analysis technology, which relate to the field of video analysis, wherein the method comprises the following steps: obtaining a first identification result of an object to be identified; rechecking the first recognition result based on an expert system to obtain a first rechecking result; rechecking the first recognition result based on an incremental learning model to obtain a second rechecking result; determining a first comprehensive rechecking result according to the first rechecking result and the second rechecking result; and evaluating the first recognition result according to the first comprehensive rechecking result. The method and the device solve the technical problem that in the prior art, the accuracy of rechecking the recognition result of the video analysis technology is not high, and further the rechecking effect is poor.
Description
Technical Field
The invention relates to the field of video analysis, in particular to a method and a system for intelligently reviewing recognition results of a video analysis technology.
Background
The video analysis technology is a high and new technology which is used for carrying out high-speed analysis on mass data in a video picture by means of the powerful calculation function of a processor, extracting key information in a video, and carrying out marking or related processing to obtain information required by a user. The video analysis technology is widely applied to the fields of national defense security, city management, cultural relic protection, financial commerce and the like. Furthermore, the scale of the recognition result of the video analysis technology is continuously enlarged, and heavy pressure is brought when the recognition result of the video analysis technology is rechecked. The research and design of the method for optimizing and rechecking the recognition result of the video analysis technology has important practical significance.
In the prior art, the technical problem that the rechecking effect is poor due to low accuracy of rechecking the recognition result of the video analysis technology exists.
Disclosure of Invention
The application provides a method and a system for intelligently rechecking a recognition result of a video analysis technology, and solves the technical problem that in the prior art, the rechecking accuracy of the recognition result of the video analysis technology is not high, and the rechecking effect is poor.
In view of the foregoing problems, the present application provides a method and system for intelligently reviewing recognition results of video analysis technologies.
In one aspect, the present application provides a method for intelligently reviewing a recognition result of a video analysis technology, where the method is applied to a system for intelligently reviewing a recognition result of a video analysis technology, and the method includes: obtaining a first identification result of an object to be identified; rechecking the first recognition result based on an expert system to obtain a first rechecking result; rechecking the first recognition result based on an incremental learning model to obtain a second rechecking result; determining a first comprehensive rechecking result according to the first rechecking result and the second rechecking result; and evaluating the first recognition result according to the first comprehensive rechecking result.
On the other hand, the application also provides a system for intelligently reviewing the recognition result of the video analysis technology, wherein the system comprises: a first obtaining unit configured to obtain a first recognition result of an object to be recognized; the second obtaining unit is used for rechecking the first recognition result based on an expert system to obtain a first rechecking result; a third obtaining unit, configured to perform review on the first recognition result based on an incremental learning model to obtain a second review result; a first execution unit, configured to determine a first comprehensive double-check result according to the first double-check result and the second double-check result; and the second execution unit is used for evaluating the first identification result according to the first comprehensive review result.
In a third aspect, the present application provides a system for intelligently reviewing recognition results of video analysis technologies, including a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor implements the steps of the method according to the first aspect when executing the program.
In a fourth aspect, the present application provides a computer-readable storage medium, wherein the storage medium has stored thereon a computer program which, when executed by a processor, implements the method of any of the first aspects described above.
One or more technical solutions provided in the present application have at least the following technical effects or advantages:
obtaining a first identification result of an object to be identified; rechecking the electronic paper by an expert system to obtain a first rechecking result; rechecking through an incremental learning model to obtain a second rechecking result; determining a first comprehensive rechecking result according to the first rechecking result and the second rechecking result; and evaluating the first recognition result according to the first comprehensive rechecking result. The method achieves the purposes of carrying out multiple rechecks on the recognition result of the video analysis technology, improving the accuracy and reliability of the rechecking, and further effectively improving the effect and quality of the rechecking; meanwhile, a method for optimizing and rechecking the recognition result of the video analysis technology is designed, and the technical effect of laying a foundation for further development of the video analysis technology is achieved.
The foregoing description is only an overview of the technical solutions of the present application, and the present application can be implemented according to the content of the description in order to make the technical means of the present application more clearly understood, and the following detailed description of the present application is given in order to make the above and other objects, features, and advantages of the present application more clearly understandable.
Drawings
In order to more clearly illustrate the technical solutions in the present application or the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only exemplary, and for those skilled in the art, other drawings can be obtained according to the provided drawings without inventive effort.
Fig. 1 is a schematic flowchart of a method for intelligently reviewing recognition results of a video analysis technique according to the present application;
fig. 2 is a schematic flow chart illustrating a process of rechecking the first recognition result based on an incremental learning model to obtain a second rechecking result in the method for intelligently rechecking the recognition result of the video analysis technology according to the present application;
fig. 3 is a schematic flowchart of a process of obtaining a second comprehensive review result according to the first review result, the second review result, and the third review result in the method for intelligently reviewing the recognition result of the video analysis technology according to the present application;
fig. 4 is a schematic structural diagram of a system for intelligently reviewing recognition results of a video analysis technique according to the present application;
fig. 5 is a schematic structural diagram of an exemplary electronic device of the present application.
Description of reference numerals: a first obtaining unit 11, a second obtaining unit 12, a third obtaining unit 13, a first executing unit 14, a second executing unit 15, an electronic device 300, a memory 301, a processor 302, a communication interface 303, and a bus architecture 304.
Detailed Description
The application provides a method and a system for intelligently rechecking a recognition result of a video analysis technology, and solves the technical problem that in the prior art, the rechecking accuracy of the recognition result of the video analysis technology is not high, and the rechecking effect is poor. The method achieves the purposes of carrying out multiple rechecks on the recognition result of the video analysis technology, improving the accuracy and reliability of the rechecking, and further effectively improving the effect and quality of the rechecking; meanwhile, a method for optimizing and rechecking the recognition result of the video analysis technology is designed, and the technical effect of laying a foundation for further development of the video analysis technology is achieved.
Hereinafter, example embodiments according to the present application will be described in detail with reference to the accompanying drawings. It should be apparent that the described embodiments are merely some embodiments of the present application and not all embodiments of the present application, and it should be understood that the present application is not limited to the example embodiments described herein.
According to the technical scheme, the data acquisition, storage, use, processing and the like meet the relevant regulations of national laws and regulations.
The video analysis technology is a high and new technology which is used for carrying out high-speed analysis on mass data in a video picture by means of the powerful calculation function of a processor, extracting key information in a video, and carrying out marking or related processing to obtain information required by a user. The video analysis technology is widely applied to the fields of national defense security, city management, cultural relic protection, financial commerce and the like. Furthermore, the scale of the recognition result of the video analysis technology is continuously enlarged, and heavy pressure is brought when the recognition result of the video analysis technology is rechecked. The research and design of the method for optimizing and rechecking the recognition result of the video analysis technology has important practical significance.
In view of the above technical problems, the technical solution provided by the present application has the following general idea:
the application provides a method for intelligently reviewing a video analysis technology recognition result, wherein the method is applied to a system for intelligently reviewing the video analysis technology recognition result, and the method comprises the following steps: obtaining a first identification result of an object to be identified; rechecking the electronic paper by an expert system to obtain a first rechecking result; rechecking through an incremental learning model to obtain a second rechecking result; determining a first comprehensive rechecking result according to the first rechecking result and the second rechecking result; and evaluating the first identification result according to the first comprehensive rechecking result.
For better understanding of the above technical solutions, the following detailed descriptions will be provided in conjunction with the drawings and the detailed description of the embodiments.
Example one
Referring to fig. 1, the present application provides a method for intelligently reviewing a recognition result of a video analysis technique, wherein the method is applied to a system for intelligently reviewing a recognition result of a video analysis technique, and the method specifically includes the following steps:
step S100: obtaining a first identification result of an object to be identified;
specifically, the object to be identified comprises any article which uses the system for intelligently reviewing the identification result of the video analysis technology to intelligently review the identification result. For example, the object to be recognized may be a license plate, a human face, clothes, food, a house, and the like. The first identification result refers to basic data information about the object to be identified, which is obtained by applying the system for intelligently rechecking the identification result of the video analysis technology to preliminarily identify the object to be identified. For example, if the object to be identified is a certain drug, the first identification result includes data information such as the production date, the manufacturer, the shelf life, the components, the applicable population, and the adverse reaction of the drug. The technical effects of defining the first recognition result of the object to be recognized and providing data support for subsequent rechecking of the object to be recognized are achieved.
Step S200: rechecking the first recognition result based on an expert system to obtain a first rechecking result;
further, step S200 of the present application further includes:
step S210: constructing an expert system based on the first recognition result, wherein the expert system comprises experts in different fields;
step S220: rechecking the first recognition result respectively through the experts in different fields;
step S230: obtaining the initial and rechecking results of the experts in different fields;
step S240: and determining the first rechecking result according to a plurality of initial rechecking results.
Specifically, on the basis of obtaining the first recognition result, an expert system is constructed; rechecking the first recognition result by using an expert system; and obtaining a plurality of initial rechecking results, and determining the first rechecking result according to the initial rechecking results. The expert system is composed of a plurality of experts in different fields and has functions of rechecking the first recognition result and the like. The expert system is included in the system for intelligently reviewing the recognition result of the video analysis technology. The initial and rechecking results are a plurality of initial and rechecking results obtained after the expert system rechecks the authenticity, correctness, reliability and the like of the first identification result. The first rechecking result is information which indicates whether the rechecking of the primary rechecking result passes or not and is obtained after comprehensive analysis and intelligent processing are carried out on a plurality of primary rechecking results by the system for intelligently rechecking the identification result of the video analysis technology. The technical effects of rechecking the first recognition result by using an expert system and improving the accuracy and reliability of the first recognition result are achieved.
Further, step S230 of the present application further includes:
step S231: judging whether the initial and rechecking results of the experts in different fields are uniform or not;
step S232: and if the initial and rechecking results of the experts in different fields are not uniform, introducing the experts in other fields to the expert system, and updating the expert pool of the expert system.
Specifically, when obtaining the initial review results of the experts in different fields, it is necessary to determine whether the initial review results of the experts in different fields in the expert system are unified; and if the initial and rechecking results of the experts in different fields are not uniform, updating the expert pool of the expert system by introducing the experts in other fields into the expert system. Wherein the expert pool of the expert system includes a talent bank comprised of a plurality of experts in different domains. Illustratively, the first identification result is data information of production date, manufacturer, shelf life, ingredients, applicable population, adverse reaction and the like of a certain medicine. When the expert system is used for rechecking the medicine, the expert in the expert system puts a question on the correctness of the component data information of the medicine, so that the initial rechecking results of the experts in different fields are not uniform; other domain experts such as pharmaceutical domain experts, medical domain experts, etc. may be introduced to the expert system and the pool of experts of the expert system updated. The technical effects of updating the expert pool of the expert system, reducing errors caused by the hysteresis quality of the expert system and the like and further improving the accuracy of the first rechecking result are achieved.
Step S300: rechecking the first recognition result based on an incremental learning model to obtain a second rechecking result;
further, as shown in fig. 2, step S300 of the present application further includes:
step S310: obtaining a recognition model of the first recognition result, wherein the recognition model is obtained by constructing a first feature and a second feature of the object to be recognized;
step S320: obtaining a third characteristic of the object to be identified;
step S330: performing incremental learning on the identification model according to the third characteristics to obtain a first identification model;
step S340: identifying the object to be identified according to the first identification model to obtain a second identification result;
step S350: and rechecking the first recognition result according to the second recognition result to obtain the second rechecking result.
Specifically, the system for intelligently rechecking the recognition result of the video analysis technology acquires the first feature, the second feature and the third feature of the object to be recognized through data mining, feature extraction and other modes. And the first characteristic, the second characteristic and the third characteristic of the object to be identified are different from each other. And training the first characteristic and the second characteristic of the object to be recognized, and further constructing a recognition model of the first recognition result. Further, incremental learning is carried out on the recognition model of the first recognition result by utilizing the third features, and a first recognition model is obtained; on the basis, the object to be identified is identified, a second identification result is obtained, the first identification result is rechecked by using the second identification result, and the second rechecking result is determined. And the second rechecking result is information which is obtained after rechecking the first recognition result by using an incremental learning model and indicates whether the first recognition result passes the rechecking. Incremental learning refers to a learning system that can continuously learn new knowledge from new samples and can preserve most of the previously learned knowledge. Incremental learning is very similar to the learning pattern of human beings themselves. Incremental learning has no need to save historical data; the occupation of storage space is reduced; the method helps users to better understand and simulate the learning mode of human brain and the composition mechanism of biological neural network from the system level, and provides technical basis for developing new calculation models and effective learning algorithms. The recognition model of the first recognition result is a neural network model formed by connecting a plurality of neurons, which is obtained by performing machine learning based on the first characteristic and the second characteristic of the object to be recognized. Therefore, the identification model is subjected to incremental learning through the third features, the obtained first identification model reserves the basic function of the identification model of the first identification result, the model updating performance is maintained continuously, and the accuracy of the second identification result is enhanced. The technical effect that the incremental learning model is used for rechecking the first recognition result so as to obtain a reliable second rechecking result and lay a foundation for subsequently determining the first comprehensive rechecking result is achieved.
Further, as shown in fig. 3, after step S350, the method further includes:
step S360: constructing a generative confrontation network based on the recognition model and the first recognition result, wherein the generative confrontation network comprises the recognition model and a discrimination model for discriminating the recognition model;
step S370: inputting the first recognition result into the discrimination model to obtain a first discrimination result;
step S380: rechecking the first recognition result according to the first judgment result to obtain a third composite result;
step S390: and obtaining a second comprehensive rechecking result according to the first rechecking result, the second rechecking result and the third rechecking result.
Specifically, on the basis of obtaining the recognition model and the first recognition result, a generative confrontation network is constructed, wherein the generative confrontation network comprises the recognition model and a discrimination model for discriminating the recognition model; inputting the first recognition result as input information into the discrimination model, and outputting a first discrimination result; rechecking the first recognition result by using the first recognition result to obtain a third composite result; further, a second composite review result is formulated based on the first review result, the second review result, and the third review result. The generative countermeasure network is an unsupervised learning method and plays a significant role in machine learning. The recognition model and a discrimination model for discriminating the recognition model are included in the generative countermeasure network. The recognition model and the discriminant model are in a confrontational relationship, the recognition model is to generate samples which cause the discriminant model to fail as much as possible, and the discriminant model is to recognize false samples of the recognition model as much as possible. The generative confrontation network continuously optimizes the identification model and the discrimination model through the confrontation relationship, and then obtains the identification model and the discrimination model with higher accuracy. The generative countermeasure network has the advantages of generating more accurate sample data, wide application range, simplicity, time saving and the like. And the third composite result is information which indicates whether the first identification result passes the rechecking after the first identification result is rechecked by using the first discrimination result obtained by the generative countermeasure network. The second comprehensive review result is a comprehensive review result obtained after the first review result, the second review result and the third review result are comprehensively analyzed by the system for intelligently reviewing the identification result of the video analysis technology. The second comprehensive rechecking result comprises two results of rechecking passing and rechecking failing. The technical effects that the first judgment result with higher accuracy is obtained by utilizing the generating type countermeasure network, the second comprehensive rechecking result with higher accuracy is further obtained, and the quality of rechecking the identification result of the video analysis technology is improved are achieved.
Step S400: determining a first comprehensive rechecking result according to the first rechecking result and the second rechecking result;
further, step S400 of the present application further includes:
step S410: if the first rechecking result and the second rechecking result both pass the rechecking, determining that the first comprehensive rechecking result passes the rechecking;
step S420: and if at least one of the first and second review results is that the review does not pass, determining that the first comprehensive review result is that the review does not pass.
Specifically, the first comprehensive review result is a comprehensive review result obtained after the system for intelligently reviewing the video analysis technology recognition result intelligently processes the first review result and the second review result. The first comprehensive rechecking result comprises two results, namely a rechecking passing result and a rechecking failing result. If the first rechecking result and the second rechecking result both are rechecked and passed, the obtained first comprehensive rechecking result is a rechecking and passing result; and if at least one of the first and second review results is that the review does not pass, obtaining the first comprehensive review result that is that the review does not pass. The technical effects of comprehensively analyzing the first rechecking result and the second rechecking result, further determining the first comprehensive rechecking result and improving the accuracy and reliability of the rechecking result are achieved.
Step S500: and evaluating the first recognition result according to the first comprehensive rechecking result.
Further, step S500 of the present application further includes:
step S510: if the first comprehensive rechecking result is passed, obtaining first label information;
step S520: marking the first identification result through the first label information;
step S530: if the first comprehensive rechecking result is failed, obtaining first reminding information;
step S540: and reminding the object to be identified of needing to be identified again through the first reminding information.
Specifically, when the first comprehensive review result is used for evaluating the first recognition result, if the first comprehensive review result passes, the system for intelligently reviewing the recognition result of the video analysis technology automatically acquires first label information for marking the first recognition result. And if the first comprehensive rechecking result is failed, automatically sending first reminding information by the system for intelligently rechecking the identification result of the video analysis technology, wherein the first reminding information is used for reminding the object to be identified again. The recognition result of the video analysis technology is rechecked for multiple times, the accuracy and the reliability of the rechecking are improved, and the rechecking effect and quality are further effectively improved; meanwhile, a method for optimizing and rechecking the recognition result of the video analysis technology is designed, and the technical effect of laying a foundation for further development of the video analysis technology is achieved.
In summary, the method for intelligently reviewing the recognition result of the video analysis technology provided by the application has the following technical effects:
1. obtaining a first identification result of an object to be identified; rechecking the electronic paper by an expert system to obtain a first rechecking result; rechecking through an incremental learning model to obtain a second rechecking result; determining a first comprehensive rechecking result according to the first rechecking result and the second rechecking result; and evaluating the first recognition result according to the first comprehensive rechecking result. The recognition result of the video analysis technology is rechecked for multiple times, the accuracy and the reliability of the rechecking are improved, and the rechecking effect and quality are further effectively improved; meanwhile, a method for optimizing and rechecking the recognition result of the video analysis technology is designed, and the technical effect of laying a foundation for further development of the video analysis technology is achieved.
2. Incremental learning refers to a learning system that can continuously learn new knowledge from new samples and can preserve most of the previously learned knowledge. Incremental learning is very similar to the learning pattern of human beings themselves. Incremental learning has no need to save historical data; the occupation of storage space is reduced; the method helps users to better understand and simulate the learning mode of human brain and the composition mechanism of biological neural network from the system level, and provides technical basis for developing new calculation models and effective learning algorithms. The recognition model of the first recognition result is a neural network model formed by connecting a plurality of neurons, which is obtained by performing machine learning based on the first characteristic and the second characteristic of the object to be recognized. Therefore, the identification model is subjected to incremental learning through the third features, the obtained first identification model reserves the basic function of the identification model of the first identification result, the model updating performance is maintained continuously, and the accuracy of the second identification result is enhanced.
3. The generative countermeasure network is an unsupervised learning method and plays a significant role in machine learning. The recognition model and a discrimination model for discriminating the recognition model are included in the generative countermeasure network. The recognition model and the discriminant model are in a confrontational relationship, the recognition model is to generate samples which cause the discriminant model to fail as much as possible, and the discriminant model is to recognize false samples of the recognition model as much as possible. The generative confrontation network continuously optimizes the identification model and the discrimination model through the confrontation relationship, and then obtains the identification model and the discrimination model with higher accuracy. The generative countermeasure network has the advantages of generating more accurate sample data, wide application range, simplicity, time saving and the like.
Example two
Based on the same inventive concept as the method for intelligently reviewing the recognition result of the video analysis technology in the foregoing embodiment, the present invention further provides a system for intelligently reviewing the recognition result of the video analysis technology, please refer to fig. 4, where the system includes:
a first obtaining unit 11, wherein the first obtaining unit 11 is used for obtaining a first identification result of an object to be identified;
a second obtaining unit 12, where the second obtaining unit 12 is configured to perform review on the first recognition result based on an expert system to obtain a first review result;
a third obtaining unit 13, where the third obtaining unit 13 is configured to perform a review on the first recognition result based on an incremental learning model to obtain a second review result;
a first execution unit 14, where the first execution unit 14 is configured to determine a first comprehensive review result according to the first review result and the second review result;
a second execution unit 15, where the second execution unit 15 is configured to evaluate the first recognition result according to the first comprehensive review result.
Further, the system further comprises:
a third execution unit, configured to construct an expert system based on the first recognition result, where the expert system includes experts in different fields;
the fourth execution unit is used for rechecking the first identification result respectively through the experts in different fields;
a fourth obtaining unit, configured to obtain a preliminary review result of the experts in the different fields;
a fifth execution unit, configured to determine the first double-check result according to a plurality of the initial double-check results.
Further, the system further comprises:
a fifth obtaining unit, configured to obtain a recognition model of the first recognition result, where the recognition model is obtained by constructing a first feature and a second feature of the object to be recognized;
a sixth obtaining unit configured to obtain a third feature of the object to be recognized;
a seventh obtaining unit, configured to perform incremental learning on the recognition model according to the third feature to obtain a first recognition model;
the eighth obtaining unit is used for identifying the object to be identified according to the first identification model to obtain a second identification result;
a ninth obtaining unit, configured to perform rechecking on the first recognition result according to the second recognition result, and obtain the second rechecking result.
Further, the system further comprises:
a sixth execution unit, configured to construct a generative confrontation network based on the recognition model and the first recognition result, where the generative confrontation network includes the recognition model and a discrimination model that discriminates the recognition model;
a tenth obtaining unit, configured to input the first recognition result into the discriminant model, and obtain a first discriminant result;
an eleventh obtaining unit, configured to perform rechecking on the first recognition result according to the first determination result to obtain a third composite result;
a twelfth obtaining unit, configured to obtain a second comprehensive review result according to the first review result, the second review result, and the third review result.
Further, the system further comprises:
a seventh execution unit, configured to determine that the first comprehensive double-check result is a double-check pass if both the first double-check result and the second double-check result are double-check passes;
an eighth execution unit, configured to determine that the first comprehensive review result is that the review does not pass if at least one of the first review result and the second review result is that the review does not pass.
Further, the system further comprises:
a thirteenth obtaining unit, configured to obtain first tag information if the first comprehensive review result is a pass;
a ninth execution unit, configured to mark the first recognition result by the first tag information;
a fourteenth obtaining unit, configured to obtain first reminding information if the first comprehensive rechecking result does not pass;
and the tenth execution unit is used for reminding the object to be identified of needing to be identified again through the first reminding information.
Further, the system further comprises:
an eleventh execution unit, configured to determine whether the initial review results of the experts in different fields are uniform;
and the twelfth execution unit is used for introducing other field experts to the expert system and updating the expert pool of the expert system if the initial and rechecking results of the experts in different fields are not uniform.
In the present specification, each embodiment is described in a progressive manner, and the main point of each embodiment is that the embodiment is different from other embodiments, and the method and the specific example for intelligently reviewing the recognition result of the video analysis technique in the first embodiment of fig. 1 are also applicable to the system for intelligently reviewing the recognition result of the video analysis technique in the present embodiment. The device disclosed by the embodiment corresponds to the method disclosed by the embodiment, so that the description is simple, and the relevant points can be referred to the method part for description.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Exemplary electronic device
The electronic device of the present application is described below with reference to fig. 5.
Based on the same inventive concept as the method for intelligently reviewing the identification result of the video analysis technology in the foregoing embodiment, the present application also provides a system for intelligently reviewing the identification result of the video analysis technology, which includes: a processor coupled to a memory, the memory for storing a program that, when executed by the processor, causes a system to perform the method of any of the first aspects.
The electronic device 300 includes: processor 302, communication interface 303, memory 301. Optionally, the electronic device 300 may also include a bus architecture 304. Wherein, the communication interface 303, the processor 302 and the memory 301 may be connected to each other through a bus architecture 304; the bus architecture 304 may be a peripheral component interconnect standard bus or an extended industry standard architecture bus, or the like. The bus architecture 304 may be divided into an address bus, a data bus, a control bus, and the like. For ease of illustration, only one thick line is shown, but this does not mean that there is only one bus or one type of bus.
The memory 301 is used for storing computer-executable instructions for executing the present application, and is controlled by the processor 302 to execute. The processor 302 is configured to execute the computer-executable instructions stored in the memory 301, so as to implement a method for intelligently reviewing recognition results of video analysis technologies provided in the present application.
Alternatively, the computer executable instructions may also be referred to as application code, and the application is not limited thereto.
The method and the device solve the technical problem that in the prior art, the accuracy of rechecking the identification result of the video analysis technology is not high, and further the rechecking effect is not good. The method achieves the purposes of carrying out multiple rechecks on the recognition result of the video analysis technology, improving the accuracy and reliability of the rechecking, and further effectively improving the effect and quality of the rechecking; meanwhile, a method for optimizing and rechecking the recognition result of the video analysis technology is designed, and the technical effect of laying a foundation for further development of the video analysis technology is achieved.
Those of ordinary skill in the art will understand that: the various numbers of the first, second, etc. mentioned in this application are for convenience of description and are not intended to limit the scope of this application nor to indicate the order of precedence. "and/or" describes the association relationship of the associated objects, meaning that there may be three relationships, e.g., a and/or B, which may mean: a exists alone, A and B exist simultaneously, and B exists alone. The character "/" generally indicates that the former and latter associated objects are in an "or" relationship. "at least one" means one or more. At least two means two or more. "at least one," "any," or similar expressions refer to any combination of these items, including any combination of singular or plural items. For example, at least one (one ) of a, b, or c, may represent: a, b, c, a-b, a-c, b-c, or a-b-c, wherein a, b, c may be single or multiple.
In the above embodiments, the implementation may be wholly or partially realized by software, hardware, firmware, or any combination thereof. When implemented in software, may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the procedures or functions described in accordance with the present application are generated, in whole or in part. The computer may be a general purpose computer, a special purpose computer, a network of computers, or other programmable device. The computer instructions may be stored in a computer readable storage medium or transmitted from one computer readable storage medium to another computer readable storage medium, for example, the computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wire or wirelessly. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device including one or more available media integrated servers, data centers, and the like. The usable medium may be a magnetic medium, an optical medium, a semiconductor medium, or the like.
The various illustrative logical units and circuits described in this application may be implemented or operated through the design of a general purpose processor, a digital signal processor, an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination thereof. A general-purpose processor may be a microprocessor, but in the alternative, the processor may be any conventional processor, controller, microcontroller, or state machine. A processor may also be implemented as a combination of computing devices, e.g., a digital signal processor and a microprocessor, a plurality of microprocessors, one or more microprocessors in conjunction with a digital signal processor core, or any other similar configuration.
The steps of a method or algorithm described in this application may be embodied directly in hardware, in a software element executed by a processor, or in a combination of the two. The software cells may be stored in RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art. For example, a storage medium may be coupled to the processor such the processor can read information from, and write information to, the storage medium. In the alternative, the storage medium may be integral to the processor. The processor and the storage medium may reside in an ASIC, which may be disposed in a terminal. In the alternative, the processor and the storage medium may reside in different components within the terminal. These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
Although the present application has been described in conjunction with specific features and embodiments thereof, it will be evident that various modifications and combinations can be made thereto without departing from the spirit and scope of the application.
Accordingly, the specification and figures are merely exemplary of the application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of the application. It will be apparent to those skilled in the art that various changes and modifications may be made in the present application without departing from the scope of the application. Thus, if such modifications and variations of the present application fall within the scope of the present application and its equivalent technology, it is intended that the present application include such modifications and variations.
Claims (10)
1. A method for intelligently reviewing recognition results of video analysis technology is characterized by comprising the following steps:
obtaining a first identification result of an object to be identified;
rechecking the first recognition result based on an expert system to obtain a first rechecking result;
rechecking the first recognition result based on an incremental learning model to obtain a second rechecking result;
determining a first comprehensive rechecking result according to the first rechecking result and the second rechecking result;
and evaluating the first recognition result according to the first comprehensive rechecking result.
2. The method of claim 1, wherein said reviewing said first recognition result based on an expert system to obtain a first review result comprises:
constructing an expert system based on the first recognition result, wherein the expert system comprises experts in different fields;
rechecking the first recognition result by the experts in different fields respectively;
obtaining the initial and rechecking results of the experts in different fields;
and determining the first rechecking result according to a plurality of initial rechecking results.
3. The method of claim 1, wherein the reviewing the first recognition result based on the incremental learning model to obtain a second review result comprises:
obtaining a recognition model of the first recognition result, wherein the recognition model is obtained by constructing a first feature and a second feature of the object to be recognized;
obtaining a third characteristic of the object to be identified;
performing incremental learning on the identification model according to the third characteristics to obtain a first identification model;
identifying the object to be identified according to the first identification model to obtain a second identification result;
and rechecking the first recognition result according to the second recognition result to obtain a second rechecking result.
4. The method of claim 3, wherein the method further comprises:
constructing a generative confrontation network based on the recognition model and the first recognition result, wherein the generative confrontation network comprises the recognition model and a discrimination model for discriminating the recognition model;
inputting the first recognition result into the discrimination model to obtain a first discrimination result;
rechecking the first recognition result according to the first judgment result to obtain a third composite result;
and obtaining a second comprehensive rechecking result according to the first rechecking result, the second rechecking result and the third rechecking result.
5. The method of claim 1, wherein said determining a first composite review result from said first review result and said second review result comprises:
if the first rechecking result and the second rechecking result both pass the rechecking, determining that the first comprehensive rechecking result passes the rechecking;
and if at least one of the first and second review results is that the review does not pass, determining that the first comprehensive review result is that the review does not pass.
6. The method of claim 1, wherein said evaluating said first recognition result based on said first composite review result comprises:
if the first comprehensive rechecking result is passed, obtaining first label information;
marking the first identification result through the first label information;
if the first comprehensive rechecking result is failed, obtaining first reminding information;
and reminding the object to be identified of needing to be identified again through the first reminding information.
7. The method of claim 2, wherein obtaining the initial review results of the plurality of domain experts further comprises:
judging whether the initial and rechecking results of the experts in different fields are uniform or not;
and if the initial and rechecking results of the experts in different fields are not uniform, introducing the experts in other fields to the expert system, and updating the expert pool of the expert system.
8. A system for intelligently reviewing recognition results of video analytics techniques, the system comprising:
a first obtaining unit configured to obtain a first recognition result of an object to be recognized;
the second obtaining unit is used for rechecking the first recognition result based on an expert system to obtain a first rechecking result;
a third obtaining unit, configured to perform review on the first recognition result based on an incremental learning model to obtain a second review result;
a first execution unit, configured to determine a first comprehensive double-check result according to the first double-check result and the second double-check result;
and the second execution unit is used for evaluating the first identification result according to the first comprehensive review result.
9. A system for intelligently reviewing the recognition results of a video analysis technique, comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the steps of the method of any one of claims 1 to 7 when executing the program.
10. A computer-readable storage medium, characterized in that the storage medium has stored thereon a computer program which, when being executed by a processor, carries out the method of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210202778.5A CN114612752A (en) | 2022-03-02 | 2022-03-02 | Method and system for intelligently reviewing recognition result of video analysis technology |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210202778.5A CN114612752A (en) | 2022-03-02 | 2022-03-02 | Method and system for intelligently reviewing recognition result of video analysis technology |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114612752A true CN114612752A (en) | 2022-06-10 |
Family
ID=81861368
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210202778.5A Pending CN114612752A (en) | 2022-03-02 | 2022-03-02 | Method and system for intelligently reviewing recognition result of video analysis technology |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114612752A (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109947651A (en) * | 2019-03-21 | 2019-06-28 | 上海智臻智能网络科技股份有限公司 | Artificial intelligence engine optimization method and device |
| JP2020004248A (en) * | 2018-06-29 | 2020-01-09 | 株式会社エヌケービー | Video reviewing device, information processing device, computer program, and video reviewing method |
| CN110826908A (en) * | 2019-11-05 | 2020-02-21 | 北京推想科技有限公司 | Evaluation method and device for artificial intelligent prediction, storage medium and electronic equipment |
| CN112327096A (en) * | 2020-10-15 | 2021-02-05 | 国家电网公司华北分部 | Fault diagnosis information fusion method and device based on self-adaptive strategy |
| WO2021082118A1 (en) * | 2019-11-01 | 2021-05-06 | 深圳大学 | Person re-identification method and apparatus, and terminal and storage medium |
| CN113095563A (en) * | 2021-04-07 | 2021-07-09 | 全球能源互联网研究院有限公司 | Method and device for reviewing prediction result of artificial intelligence model |
| CN114049895A (en) * | 2022-01-12 | 2022-02-15 | 北京荣达天下信息科技有限公司 | ASR-based voice quality inspection analysis method and system |
-
2022
- 2022-03-02 CN CN202210202778.5A patent/CN114612752A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2020004248A (en) * | 2018-06-29 | 2020-01-09 | 株式会社エヌケービー | Video reviewing device, information processing device, computer program, and video reviewing method |
| CN109947651A (en) * | 2019-03-21 | 2019-06-28 | 上海智臻智能网络科技股份有限公司 | Artificial intelligence engine optimization method and device |
| WO2021082118A1 (en) * | 2019-11-01 | 2021-05-06 | 深圳大学 | Person re-identification method and apparatus, and terminal and storage medium |
| CN110826908A (en) * | 2019-11-05 | 2020-02-21 | 北京推想科技有限公司 | Evaluation method and device for artificial intelligent prediction, storage medium and electronic equipment |
| CN112327096A (en) * | 2020-10-15 | 2021-02-05 | 国家电网公司华北分部 | Fault diagnosis information fusion method and device based on self-adaptive strategy |
| CN113095563A (en) * | 2021-04-07 | 2021-07-09 | 全球能源互联网研究院有限公司 | Method and device for reviewing prediction result of artificial intelligence model |
| CN114049895A (en) * | 2022-01-12 | 2022-02-15 | 北京荣达天下信息科技有限公司 | ASR-based voice quality inspection analysis method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Benchaji et al. | Enhanced credit card fraud detection based on attention mechanism and LSTM deep model | |
| Yang et al. | Reconstruction of complex directional networks with group lasso nonlinear conditional granger causality | |
| WO2022105115A1 (en) | Question and answer pair matching method and apparatus, electronic device and storage medium | |
| WO2017132545A1 (en) | Systems and methods for generative learning | |
| Tattar | Hands-On Ensemble Learning with R: A beginner's guide to combining the power of machine learning algorithms using ensemble techniques | |
| CN106537423A (en) | Adaptive featurization as service | |
| García-Magariño et al. | Estimation of missing prices in real-estate market agent-based simulations with machine learning and dimensionality reduction methods | |
| CN110188158A (en) | Keyword and topic label generating method, device, medium and electronic equipment | |
| CN110458644A (en) | A kind of information processing method and relevant device | |
| CN115170319A (en) | Abnormal account detection method, and method and device for constructing graph neural network model | |
| Zhang et al. | Cell type matching in single-cell RNA-sequencing data using FR-Match | |
| CN115455382A (en) | Semantic comparison method and device for binary function codes | |
| Auffarth | Artificial Intelligence with Python Cookbook: Proven recipes for applying AI algorithms and deep learning techniques using TensorFlow 2. x and PyTorch 1.6 | |
| CN110288488A (en) | Medical insurance Fraud Prediction method, apparatus, equipment and readable storage medium storing program for executing | |
| CN109885745A (en) | A kind of user draws a portrait method, apparatus, readable storage medium storing program for executing and terminal device | |
| CN105335630A (en) | Identity recognition method and identity recognition device | |
| CN114612752A (en) | Method and system for intelligently reviewing recognition result of video analysis technology | |
| US20220092448A1 (en) | Method and system for providing annotation information for target data through hint-based machine learning model | |
| CN114611592A (en) | Semi-supervised feature selection method, system, medium, equipment and terminal | |
| Asilkalkan et al. | Matrix‐variate time series modelling with hidden Markov models | |
| CN116092094A (en) | Image text recognition method and device, computer readable medium and electronic equipment | |
| CN113420847A (en) | Target object matching method based on artificial intelligence and related equipment | |
| CN112149743A (en) | Access control method, device, equipment and medium | |
| Liu et al. | Time series data augmentation method of small sample based on optimized generative adversarial network | |
| Sánchez-Gutiérrez et al. | Addressing the class imbalance in tabular datasets from a generative adversarial network approach in supervised machine learning |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |