CN117314269A - Quality inspection model management method, system, electronic equipment and storage medium - Google Patents

Abstract

The embodiment of the application provides a quality inspection model management method, a quality inspection model management system, electronic equipment and a storage medium, and belongs to the technical field of model management. The method comprises the following steps: acquiring an initial quality inspection model, and performing quality inspection operation on a target object according to the initial quality inspection model to obtain initial quality inspection parameters; when the initial quality inspection parameters are monitored to be not in the preset quality inspection parameter range, screening from a database, obtaining sample training data, and training an initial quality inspection model according to the sample training data; screening and obtaining sample test data from a database, testing the trained initial quality inspection model according to the sample test data, and obtaining model test parameters; when the model test parameters are matched with the preset model standard reaching values, determining the initial quality inspection model after the test as a target quality inspection model, wherein the target quality inspection model is used for carrying out quality inspection operation on the target object again. The quality inspection model updating method and device can reduce the updating difficulty of the quality inspection model and improve the updating efficiency of the quality inspection model.

Description

Quality inspection model management method, system, electronic equipment and storage medium

Technical field

The present application relates to the field of model management technology, and in particular to a quality inspection model management method, system, electronic equipment and storage medium.

Background technique

At present, quality inspection models are often used to improve the processing efficiency of quality inspection work. For example, in industrial quality inspection, industrial quality inspection models are often used to analyze and process relevant data of a large number of industrial products. Moreover, the use of industrial quality inspection models has improved the efficiency of quality inspection work. Quality inspection efficiency of industrial products.

However, in the actual application of industrial quality inspection models, quality inspection output results will not match the changing quality inspection requirements over time, which ultimately leads to the industrial quality inspection model not meeting the accuracy requirements of industrial quality inspection. In related technologies, the adjustment and update of industrial quality inspection models rely on the development and iteration of professional engineers. However, the actual users of industrial quality inspection models are usually factory personnel. The traditional iteration method of updating industrial quality inspection models has many steps and is professional. Problems such as strong and long update time make the update complexity of the industrial quality inspection model high, resulting in low update efficiency of the industrial quality inspection model, which in turn affects the efficiency of industrial quality inspection related work.

Contents of the invention

The main purpose of the embodiments of this application is to propose a quality inspection model management method, system, electronic device and storage medium, which can reduce the difficulty of updating the quality inspection model and improve the update efficiency of the quality inspection model.

In order to achieve the above purpose, the first aspect of the embodiment of the present application proposes a quality inspection model management method. The method includes: obtaining an initial quality inspection model, performing quality inspection operations on the target object according to the initial quality inspection model, and obtaining Initial quality inspection parameters; when monitoring that the initial quality inspection parameters are not within the preset quality inspection parameter range, filter and obtain sample training data from the database, and train the initial quality inspection model based on the sample training data ; Screen and obtain sample test data from the database, test the initial quality inspection model after training according to the sample test data, and obtain model test parameters; when the model test parameters match the preset model compliance value When, the initial quality inspection model after testing is determined to be the target quality inspection model, and the target quality inspection model is used to re-perform quality inspection operations on the target object.

In some embodiments, after testing the trained initial quality inspection model based on the sample test data and obtaining model test parameters, the method further includes: when the model test parameters meet the preset model standards When the values do not match, re-screen and obtain sample training data and sample test data from the database; re-train and update the initial quality inspection model based on the re-determined sample training data and sample test data. , and obtain the updated model test parameters. When the model test parameters match the preset model compliance value, the updated initial quality inspection model is determined to be the target quality inspection model.

In some embodiments, when it is monitored that the initial quality inspection parameters are not within the preset quality inspection parameter range, filtering and obtaining sample training data from the database includes: parsing the initial quality inspection parameters to obtain the A plurality of parameter categories of the initial quality inspection parameters, each parameter category includes a corresponding confidence level, the confidence level is used to represent the degree of credibility of the initial quality inspection model belonging to the parameter category; when the confidence level When the confidence level is not within the preset quality inspection parameter range, the screening category of the sample training data screened from the database is determined according to the confidence level, and the corresponding sample training data is selected from the database according to the screening category.

In some embodiments, the sample training data includes difficult example data and regular data; training the initial quality inspection model based on the sample training data includes: when determining the sample training data from a database , activate the training trigger of the initial quality inspection model; according to the training trigger in the activated state, input the difficult case data into the initial quality inspection model, obtain the difficult case training results, and put the difficult case data into the initial quality inspection model. Regular data is input into the initial quality inspection model to obtain regular training results; based on the difficult example training results and the regular training results, training of the initial quality inspection model is completed.

In some embodiments, filtering and obtaining sample test data from the database, testing the initial quality inspection model after training according to the sample test data, and obtaining model test parameters includes: when completing the When training the initial quality inspection model, activate the test trigger of the initial quality inspection model; filter and obtain sample test data from the database according to the test trigger in the activated state; perform training based on the sample test data The subsequent initial quality inspection model is tested and the model test parameters are obtained.

In some embodiments, after determining that the trained initial quality inspection model is the target quality inspection model, the method further includes: displaying a quality inspection model management interface, and displaying the initial quality inspection model on the quality inspection model management interface. A configuration component, wherein the configuration component includes a data configuration component, a model training component, a model testing component and a model publishing component; in response to a click operation on the data configuration component, the configured sample data is displayed on the quality inspection model management interface, Wherein, the sample data includes sample training data and sample test data; in response to the click operation of the model training component, model training parameter values are displayed on the quality inspection model management interface, and the model training parameter values are based on the initial quality inspection model. The sample training data is obtained after training; in response to the click operation of the model test component, the model test parameter value is displayed on the quality inspection model management interface, and the model test parameter value is the initial quality inspection model according to the sample The test data is obtained after testing, and the initial quality inspection model after passing the test is determined to be the target quality inspection model; in response to the click operation of the model release component, the released target quality inspection model is displayed on the quality inspection model management interface .

In some embodiments, displaying the configured sample data on the quality inspection model management interface includes: displaying a quality inspection model configuration interface, displaying a data generation component of the sample data on the quality inspection model configuration interface, and Data annotation component; in response to the click operation of the data generation component, the sample training data and sample test data obtained according to the initial quality inspection parameter configuration are displayed in the data configuration area of the quality inspection model configuration interface, wherein the sample training data Including difficult example data and regular data; in response to the click operation of the data annotation component, the annotation results for the sample training data are displayed in the data annotation area of the quality inspection model configuration interface, wherein the annotation results include the difficult example data annotation results and conventional data annotation results; in response to the completion of the data configuration operation, the configured sample data is displayed on the quality inspection model management interface according to the sample test data and the sample training data containing the annotation results.

In order to achieve the above purpose, the second aspect of the embodiment of the present application proposes a quality inspection model management system. The system includes: an acquisition module, used to obtain an initial quality inspection model, and perform inspection on the target object according to the initial quality inspection model. The quality inspection operation is to obtain initial quality inspection parameters; the model training module is used to filter and obtain sample training data from the database when it is monitored that the initial quality inspection parameters are not within the preset quality inspection parameter range. The training data trains the initial quality inspection model; the model testing module is used to screen and obtain sample test data from the database, test the trained initial quality inspection model according to the sample test data, and obtain the model Test parameters; target quality inspection model module, used to determine that the initial quality inspection model after testing is the target quality inspection model when the model test parameters match the preset model compliance value, and the target quality inspection model is used To re-perform the quality inspection operation on the target object.

In order to achieve the above object, a third aspect of the embodiment of the present application proposes an electronic device. The electronic device includes a memory and a processor. The memory stores a computer program. When the processor executes the computer program, the above is implemented. The method described in the embodiment of the first aspect.

In order to achieve the above object, the fourth aspect of the embodiment of the present application proposes a storage medium. The storage medium is a computer-readable storage medium. The storage medium stores a computer program. The computer program is implemented when executed by a processor. The method described in the above embodiment of the first aspect.

The quality inspection model management method, system, electronic equipment and storage medium proposed in this application first obtain the initial quality inspection model and perform quality inspection operations on the target object according to the initial quality inspection model to obtain the initial quality inspection parameters; then, when When it is monitored that the initial quality inspection parameters are not within the preset quality inspection parameter range, the sample training data is screened from the database and the initial quality inspection model is trained based on the sample training data. In this way, the initial quality inspection can be performed based on the sample training data. The model is trained to adapt to the changing quality inspection needs; then, sample test data is screened from the database and sample test data is obtained. The trained initial quality inspection model is tested based on the sample test data and the model test parameters are obtained; when the model test parameters When it matches the preset model compliance value, the initial quality inspection model after testing is determined to be the target quality inspection model, and the target quality inspection model is used to re-perform quality inspection operations on the target object. It can be understood that the initial quality inspection model after passing the test means that it can meet the current quality inspection needs after training and testing. Therefore, the initial quality inspection model after passing the test can be determined as the target quality inspection model, and use The target quality inspection model re-inspects the target object. In the actual quality inspection model update process, only the sample training data and sample test data are obtained from the preset database to automatically execute the iteration of the quality inspection model. Update to reduce the difficulty of updating the quality inspection model, improve the update efficiency of the quality inspection model, and thereby improve the efficiency of quality inspection work.

Description of drawings

Figure 1 is a schematic diagram of the application scenario of the quality inspection model management system provided by the embodiment of the present application;

Figure 2 is an optional flow chart of the quality inspection model management method provided by the embodiment of the present application;

Figure 3 is an implementation flow chart after step S104 in Figure 2;

Figure 4 is an implementation flow chart of step S102 in Figure 2;

Figure 5 is another implementation flow chart of step S102 in Figure 2;

Figure 6 is an implementation flow chart of step S103 in Figure 2;

Figure 7 is a schematic diagram of an optional workflow configuration of the quality inspection model management method provided by the embodiment of the present application;

Figure 8 is a schematic diagram of another optional workflow configuration of the quality inspection model management method provided by the embodiment of the present application;

Figure 9 is an optional workflow execution diagram of quality inspection model management provided by the embodiment of the present application;

Figure 10 is an optional flow diagram of the quality inspection model management method provided by the embodiment of the present application;

Figure 11 is another optional flow chart of the quality inspection model management method provided by the embodiment of the present application;

Figure 12 is an implementation flow chart of step S602 in Figure 11;

Figure 13 is a schematic diagram of the functional modules of quality inspection model management provided by the embodiment of the present application;

Figure 14 is a schematic diagram of the hardware structure of an electronic device provided by an embodiment of the present application.

Detailed ways

In order to make the purpose, technical solutions and advantages of the present application more clear, the present application will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described here are only used to explain the present application and are not used to limit the present application.

It should be noted that although the functional modules are divided in the device schematic diagram and the logical sequence is shown in the flow chart, in some cases, the modules can be divided into different modules in the device or the order in the flow chart can be executed. The steps shown or described. The terms "first", "second", etc. in the description, claims, and above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific sequence or sequence.

Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the technical field to which this application belongs. The terms used herein are only for the purpose of describing the embodiments of the present application and are not intended to limit the present application.

First, let’s analyze some terms involved in this application:

Artificial intelligence (AI): It is a new technical science that studies and develops theories, methods, technologies and application systems for simulating, extending and expanding human intelligence; artificial intelligence is a branch of computer science, artificial intelligence Intelligence attempts to understand the essence of intelligence and produce a new intelligent machine that can respond in a manner similar to human intelligence. Research in this field includes robotics, language recognition, image recognition, natural language processing, and expert systems. Artificial intelligence can simulate the information process of human consciousness and thinking. Artificial intelligence is also a theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

Machine Learning Operations (MLOps) refers to applying the best practices of software engineering to the development, deployment and maintenance process of machine learning models to improve the reliability, scalability and repeatability of machine learning systems. Nature is continuous delivery and automated pipelines in machine learning.

Data extraction, transformation and loading (ETL) is a data integration process, and its name stands for "Extract", "Transform", and "Load". ETL is typically used to extract data from disparate data sources (such as databases, files, or APIs), then transform it and finally load it into a target data storage location (such as a data warehouse or data lake).

Message Queue (MQ) is used for message delivery between different services.

Workflow is a specific pipeline run by MLOps. A workflow is divided into many steps, and these steps have dependencies. If you follow the steps of the workflow and execute them step by step, an update iteration of the model will be completed.

Workflow server is a server software used to manage and execute workflow. It provides a way to centrally manage and automate the execution of various business processes, helping organizations improve work efficiency, optimize business processes, and enable task collaboration and monitoring.

Based on this, embodiments of the present application provide a quality inspection model management method, system, electronic device and storage medium, which can reduce the difficulty of updating the quality inspection model and improve the update efficiency of the quality inspection model.

The quality inspection model management method, system, electronic device and storage medium provided by the embodiment of the present application are specifically described through the following embodiments. First, the quality inspection model management system in the embodiment of the present application is described.

At present, quality inspection models are often used to improve the processing efficiency of quality inspection work. For example, in industrial quality inspection, industrial quality inspection models are often used to analyze and process relevant data of a large number of industrial products. Moreover, the use of industrial quality inspection models has improved the efficiency of quality inspection work. Quality inspection efficiency of industrial products.

However, in the actual application of industrial quality inspection models, quality inspection output results will not match the changing quality inspection requirements over time, which ultimately leads to the industrial quality inspection model not meeting the accuracy requirements of industrial quality inspection. In related technologies, the adjustment and update of industrial quality inspection models rely on the development and iteration of professional engineers. However, the actual users of industrial quality inspection models are usually factory personnel. The traditional iteration method of updating industrial quality inspection models has many steps and professionalism. Problems such as strong and long update time make the update complexity of the industrial quality inspection model high, resulting in low update efficiency of the industrial quality inspection model, which in turn affects the efficiency of industrial quality inspection related work.

Based on this, the quality inspection model management method, system, electronic device and storage medium proposed in this application first obtain the initial quality inspection model and perform quality inspection operations on the target object according to the initial quality inspection model to obtain the initial quality inspection parameters; Then, when it is monitored that the initial quality inspection parameters are not within the preset quality inspection parameter range, the sample training data is screened from the database and the initial quality inspection model is trained based on the sample training data. In this way, the sample training data can be used to The initial quality inspection model is trained to adapt to the changing quality inspection needs; then, the sample test data is screened from the database, and the trained initial quality inspection model is tested based on the sample test data, and the model test parameters are obtained; when When the model test parameters match the preset model compliance value, the initial quality inspection model after testing is determined to be the target quality inspection model, and the target quality inspection model is used to re-perform quality inspection operations on the target object. It can be understood that the initial quality inspection model after passing the test means that it can meet the current quality inspection needs after training and testing. Therefore, the initial quality inspection model after passing the test can be determined as the target quality inspection model, and use The target quality inspection model re-inspects the target object. In the actual quality inspection model update process, only the sample training data and sample test data are obtained from the preset database to automatically execute the iteration of the quality inspection model. Update to reduce the difficulty of updating the quality inspection model, improve the update efficiency of the quality inspection model, and thereby improve the efficiency of quality inspection work.

The quality inspection model management method, system, electronic device and storage medium provided by the embodiments of this application are specifically explained through the following embodiments.

Illustratively, as shown in Figure 1, Figure 1 is a schematic diagram of the application scenario of the quality inspection model management system provided by the embodiment of the present application. The quality inspection model management system (hereinafter referred to as "system") may include a user terminal 11 and Management terminal 12, in which the quality inspection model management method and the quality inspection model management system of the embodiment of the present application can both be run on the user terminal 11, or the quality inspection model management system provided by the embodiment of the present application is run on the user terminal 11, and Run the quality inspection model management method provided by the embodiment of the present application on the management end 12, connect the user end 11 and the management end 12 through remote connection, and apply the quality inspection model management method in the management end 12 through remote control. In the quality inspection model management system of client 11. The settings are specifically set according to the actual needs of the operator, and are not specifically limited in the embodiments of this application.

The quality inspection model management method in the embodiment of the present application can be explained through the following embodiment.

In some embodiments, the embodiments of the present application can also acquire and process relevant data based on artificial intelligence technology. For example, artificial intelligence can be used to intelligently divide the acquired data to obtain sample difficult case data and sample regular data, and store the sample difficult case data and sample regular data in the database as sample training data, reducing the need for artificial processing of large amounts of data. The screening and division further improves the update efficiency of the quality inspection model. Among them, artificial intelligence (AI) is the theory, method, technology and application system that uses digital computers or digital computer-controlled machines to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results. .

Basic artificial intelligence technologies generally include technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technology, operation/interaction systems, mechatronics and other technologies. Artificial intelligence software technology mainly includes computer vision technology, robotics technology, biometric technology, speech processing technology, natural language processing technology, and machine learning/deep learning.

The quality inspection model management method provided by the embodiment of this application relates to the field of artificial intelligence technology. The quality inspection model management method provided by the embodiment of the present application can be applied in a terminal or a server, or can be software running in a terminal or a server. In some embodiments, the terminal can be a smartphone, a tablet, a laptop, a desktop computer, etc.; the server can be configured as an independent physical server, or as a server cluster or distributed system composed of multiple physical servers. A cloud that can be configured to provide basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms. Server; software can be an application that implements quality inspection model management methods, etc., but is not limited to the above forms.

The application may be used in a variety of general or special purpose computer system environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics devices, network PCs, minicomputers, mainframe computers, including Distributed computing environment for any of the above systems or devices, etc. The application may be described in the general context of computer-executable instructions, such as program modules, being executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. The present application may also be practiced in distributed computing environments where tasks are performed by remote processing devices connected through a communications network. In a distributed computing environment, program modules may be located in both local and remote computer storage media including storage devices.

It should be noted that in each specific implementation of the present application, when it comes to relevant processing based on user information, user behavior data, user historical data, user location information and other data related to user identity or characteristics, the first step is to perform relevant processing. Obtain the user's permission or consent. Moreover, the collection, use and processing of this data will comply with relevant laws, regulations and standards. In addition, when the embodiment of this application needs to obtain the user's sensitive personal information, it will obtain the user's separate permission or separate consent through a pop-up window or jump to a confirmation page. After clearly obtaining the user's separate permission or separate consent, it will then Obtain necessary user-related data for normal operation of the embodiment of the present application.

Based on this, the quality inspection model management method in the embodiment of the present application can be explained through the following embodiment.

As shown in Figure 2, Figure 2 is an optional flow chart of the quality inspection model management method provided by the embodiment of the present application. The method in Figure 2 may include, but is not limited to, steps S101 to S104.

Step S101, obtain an initial quality inspection model, perform quality inspection operations on the target object according to the initial quality inspection model, and obtain initial quality inspection parameters;

In some embodiments, the initial quality inspection model can be applied in the industrial field, and the target object can be a product or component in industrial manufacturing, for example, it can be an automobile component in automobile industry production.

In some embodiments, the initial quality inspection model may be generated based on initial quality inspection requirements. For example, a certain batch of automobile engine parts needs to be quality inspected, that is, the initial quality inspection model is generated based on the first quality inspection requirements. However, As time goes by, a second quality inspection requirement arises. At this time, if the initial quality inspection model generated by the first quality inspection requirement is used to inspect the automobile engine parts, the expected quality inspection results cannot be achieved. , therefore, the initial quality inspection model needs to be updated.

In some embodiments, the quality inspection of the target object may include quality inspection of appearance defects, such as cracks, damage, and deformation, and may also include quality inspection of dimensional requirements, such as diameter, length, and flatness.

In some embodiments, the initial quality inspection parameters can be obtained by performing quality inspection operations on the target object using the initial quality inspection model. For example, when quality inspection is performed on the dimensions of automobile engine parts, the initial quality inspection parameters are the dimensions of the automobile engine parts. The data records, in particular, may also include diameter data, length data and flatness data.

It should be noted that the initial quality inspection model can also be used in other product manufacturing industries that require quality inspection operations, and the target object can also be adjusted according to the actual situation. The embodiments of this application are only explained with preferred embodiments and are not Make specific restrictions.

In some embodiments, the iterative update of the initial quality inspection model can be performed on a platform equipped with MLOps capabilities (hereinafter referred to as the "platform"), such as Kubeflow, MLflow, TensorFlow Extended, or DataRobot. Because platforms equipped with MLOps capabilities usually have a series of built-in tools and components suitable for machine learning development and management, such as model version control systems, automated deployment tools, model performance monitoring systems, etc., these tools and components can make developers more convenient The development, deployment and management of models can be carried out independently. By performing the update of the initial quality inspection model on a platform equipped with MLOps capabilities, the efficiency of the update and iteration of the quality inspection model can be improved.

Step S102, when it is monitored that the initial quality inspection parameters are not within the preset quality inspection parameter range, sample training data is obtained from the database, and the initial quality inspection model is trained based on the sample training data;

In some embodiments, the preset quality inspection parameter range is actually the quality inspection requirement. The quality inspection parameter range can be adjusted according to actual needs, and the preset quality inspection parameter range can be input into a platform equipped with MLOps capabilities. . Among them, the preset quality inspection parameter range can be updated regularly by relevant operators.

In some embodiments, it is necessary to filter the database according to the category of the initial quality inspection parameter to obtain corresponding sample training data. For example, when using the initial quality inspection model to conduct dimensional quality inspection of automobile engine parts, the output is the diameter quality inspection result. At this time, it is necessary to filter the data of the "size-diameter" category in the database to obtain Sample training data for training the initial quality inspection model.

In some embodiments, a database may be deployed in the platform, or sample data may be obtained through a communication connection to a server where the database is deployed, where the sample data includes sample training data and sample test data.

In some embodiments, the platform may include a controller for real-time monitoring of the generated initial quality inspection parameters. When the initial quality inspection parameters are not within the preset quality inspection parameter range, the acquisition of sample training data may be automatically triggered, or , a prompt alarm can be issued to allow relevant operators to confirm whether the initial quality inspection model needs to be updated.

In some embodiments, the data in the database can be collected and updated regularly. For example, relevant operators can regularly update the data in the database based on experience, or can obtain it from a designated data source through an update script.

In some embodiments, the initial quality inspection model can be trained separately by acquiring multiple sets of sample training data to obtain multiple different trained initial quality inspection models.

In some embodiments, after completing the training of the initial quality inspection model, a preliminary evaluation of the trained initial quality inspection model can also be performed. Commonly used evaluation methods include using confusion matrix, characteristic curve (Receiver Operating Characteristic Curve, ROC) and Area under the curve (AUC) is used to evaluate the characteristics, and the evaluation metrics usually include accuracy, precision, recall, F1 score, etc.

Step S103, filter and obtain sample test data from the database, test the trained initial quality inspection model based on the sample test data, and obtain model test parameters;

In some embodiments, after completing the training of the initial quality inspection model, the trained initial quality inspection model needs to be tested and the model test parameters can be obtained to determine whether the trained initial quality inspection model meets the expected quality inspection. Require.

In some embodiments, if when training the initial quality inspection model, different training versions of the initial quality inspection model under multiple sets of sample training data are generated, the multiple sets of different initial quality inspection models obtained after training can be tested separately. .

In some embodiments, testing the trained initial quality inspection model may include A/B testing and online indicator observation. Among them, A/B testing refers to comparing the initial quality inspection models after training to determine which initial quality inspection model after training performs better. Online indicator observation refers to running the initial quality inspection model under A/B testing, and comparing the indicators between A/B tests, such as model accuracy, recall rate, etc., and comprehensively judging which version is available through online real-time operation. The model is better.

In some embodiments, in order to ensure the authenticity of the sample test data, the sample test data obtained from the database can be randomly generated. In this way, the sample test data can be made more representative to enhance the robustness of the quality inspection model. sex and stability.

Step S104: When the model test parameters match the preset model compliance value, the initial quality inspection model after testing is determined to be the target quality inspection model, and the target quality inspection model is used to re-perform quality inspection operations on the target object.

In some embodiments, a model compliance value is preset, where the model compliance value is used to indicate whether the updated initial quality inspection model achieves the expected quality inspection effect.

In some embodiments, the model test parameters obtained from the initial quality inspection model test can be compared with the preset model compliance values. If the model test parameters fall within the preset model compliance value range, it means that the initial model test parameters after training and testing If the quality inspection model meets the expected value of the model, the updated initial quality inspection model is determined to be the target quality inspection model. Use the target quality inspection model to perform a new quality inspection operation on the target object to complete accurate quality inspection of the target object.

It should be noted that during the process of updating the initial quality inspection model, the decline in quality inspection accuracy of the initial quality inspection model can be automatically monitored by the platform, and the corresponding sample training data and sample test data are also pre-configured in the database. It does not require professionals to manually mine and filter data. After testing, the quality inspection model with the best test effect is selected as the target quality inspection model. It is understandable that even factory operators who are not very professional can complete the update of the quality inspection model independently, which improves the efficiency of the update of the quality inspection model. Moreover, the updated and determined target quality inspection model can be immediately applied to the quality inspection of the target object. This greatly improves the efficiency of the quality inspection work. If the problem of quality inspection accuracy decline occurs again in actual use, it can still be used. Update the quality inspection model again.

As shown in Figure 3, Figure 3 is an implementation flow chart after step S104 in Figure 2. In some embodiments, steps S201 to S202 may also be included after step S104:

Step S201, when the model test parameters do not match the preset model compliance value, re-screen the database and obtain sample training data and sample test data;

In some embodiments, if the model test parameters do not match the preset model compliance value, it means that the training of the initial quality inspection model has not achieved the expected results, and it is necessary to re-screen the database and obtain sample training data and sample test data to correct the problem. The initial quality inspection model is trained and tested again. Among them, the sample training data screened during retraining should be different from the sample training data used in previous training, and the sample test data can be the same data to better detect the training effect of the initial quality inspection model after training.

In some embodiments, when the model test parameters do not match the preset model compliance value, the sample training data can also be filtered based on the deviation value between the model test parameters and the preset model compliance value. For example, when the model test parameters When the deviation value range from the preset model standard value is small, it means that the updated quality inspection model may have an overfitting problem. At this time, the first training data is selected, and the first training data includes diverse sample training data; When the deviation range between the model test parameters and the preset model compliance value is large, it means that the quality of the updated quality inspection model may not be high. At this time, the second training data is selected, and the second training data includes high-quality samples. training data, so that the initial quality inspection model can be trained more specifically.

Step S202: Based on the redetermined sample training data and sample test data, re-train and update the initial quality inspection model and obtain updated model test parameters. When the model test parameters match the preset model compliance value , determine the updated initial quality inspection model as the target quality inspection model

In some embodiments, after the sample training data and sample test data are re-determined, the sample training data and sample test data can be used to retrain and test the initial quality inspection model. If the updated model obtained after retraining and testing is tested If the parameters still do not match the preset model compliance value, steps S201 to S202 need to be repeatedly executed until the model test parameters match the preset model compliance value, and the target quality inspection model is obtained.

As shown in Figure 4, Figure 4 is an implementation flow chart of step S102 in Figure 2. In some embodiments, step S102 may include steps S301 to step S302:

Step S301: Analyze the initial quality inspection parameters to obtain multiple parameter categories of the initial quality inspection parameters. Each parameter category includes a corresponding confidence level. The confidence level is used to represent the degree of credibility of the initial quality inspection model belonging to the parameter category;

In some embodiments, when the initial quality inspection parameters are not within the preset quality inspection parameter range, the initial quality inspection parameters can be parsed. The purpose of the analysis is to obtain the parameter category of the initial quality inspection parameters. For example, when the initial quality inspection model is an image classification model, a result list will be output. The result list includes the parameter category of each quality inspection target object and the confidence level corresponding to the parameter category.

In some embodiments, if the obtained initial quality inspection parameter is a parameter code, the code can be parsed and the corresponding parameter category can be obtained.

Step S302: When the confidence level is not within the preset quality inspection parameter range, determine the screening category of the sample training data screened from the database based on the confidence level, and select the corresponding sample training data from the database based on the screening category.

In some embodiments, when the confidence obtained by analysis is not within the preset quality inspection parameter range, it can be determined that the quality inspection accuracy of the initial quality inspection model has declined, and the initial quality inspection model needs to be trained and updated, and the initial quality inspection model needs to be updated from the database. The basis for selecting sample training data is the initial quality inspection parameters. Specifically, the sample training data screened from the database is determined according to the parameter category, and the sample number of the screened sample training data is determined based on the confidence level.

In some embodiments, the sample training data includes difficult case data, where the difficult case data can be filtered through SQL or ETL, or a screening program can be provided, and which strategy to use is related to the user's specific scenario.

As shown in Figure 5, Figure 5 is another implementation flow chart of step S102 in Figure 2. In some embodiments, step S102 may also include steps S401 to step S403:

Step S401, when the sample training data is determined from the database, activate the training trigger of the initial quality inspection model;

In some embodiments, the sample training data includes hard case data and regular data, where the regular data is data that the initial quality inspection model often encounters during the actual quality inspection process, and the hard case data is the data that the initial quality inspection model often encounters during the actual quality inspection process. data that is not commonly encountered in .

In some embodiments, a training trigger (Trigger) is also provided. When the sample training data is determined from the database, the training trigger of the initial quality inspection model can be activated.

Step S402, according to the training trigger in the activated state, input the difficult case data into the initial quality inspection model to obtain the difficult case training results, and input the regular data into the initial quality inspection model to obtain the regular training results;

In some embodiments, when the training trigger is in the activated state, it can be determined that the initial quality inspection model needs to be trained based on the obtained sample training data. First, the difficult example data can be input into the initial quality inspection model, because it is difficult to Example data usually represents data under extreme or boundary conditions and plays a key role in the training of the initial quality inspection model, which can improve the accuracy and generalization ability of the model.

In some embodiments, hard case data may include boundary data, special format data, abnormal data, extreme data, etc.

In some embodiments, after completing the test on the difficult example data, it is usually necessary to test the initial quality inspection model based on regular data to observe the general data processing capabilities of the trained initial quality inspection model.

In some embodiments, after inputting regular data into the initial quality inspection model, the initial quality inspection model can be trained based on the input regular data and obtain regular training results. Generally, the hard case training results are similar to the regular training results. Including two situations: correct training results and wrong training results, as well as difficult example training results and regular training results, it can be decided whether it is necessary to increase sample training data.

Step S403: Complete the training of the initial quality inspection model based on the difficult example training results and the regular training results.

In some embodiments, the difficult example data may be a difficult example data set. Training the initial quality inspection model based on the difficult example data set can obtain multiple difficult example training results, and the target percentage of the difficult example training results can be set. If the accuracy rate of setting the difficult example data set reaches 70%, the difficult example training is passed.

In some embodiments, if the difficult example training fails, there is no need to perform training on regular data, because the initial quality inspection model represents its quality inspection of the target object corresponding to the difficult example data without meeting the requirements of extreme data training. There are still quality inspection errors, and in quality inspection, especially industrial quality inspection activities, the training of difficult example data is very critical. Therefore, if the difficult example training fails, the sample training data and sample training can be directly re-obtained from the database data, there is no need to perform training on the current sample training data. In this way, the update efficiency of the quality inspection model can be further improved.

As shown in Figure 6, Figure 6 is an implementation flow chart of step S103 in Figure 2. In some embodiments, step S103 may also include steps S501 to S503:

Step S501, when the training of the initial quality inspection model is completed, activate the test trigger of the initial quality inspection model;

In some embodiments, a test trigger is provided, and when the sample training data is determined from the database, the test trigger of the initial quality inspection model can be activated.

Step S502, filter and obtain sample test data from the database according to the activated test trigger;

In some embodiments, when the test trigger is in the activated state, it can be determined that the initial quality inspection model needs to be tested. Similar to the training process of the initial quality inspection model, the sample test data of the initial quality inspection model needs to be determined from the database first. Through the test trigger, the acquisition of sample test data is triggered, which avoids the tedious steps of manually waiting and monitoring the training of the initial quality inspection model and manually starting the initial quality inspection model test during traditional quality inspection model updates. Further, Improved the efficiency of updating quality inspection models.

In some embodiments, the sample test data can also be input in real time by relevant operators, and can be set according to the actual situation. This application does not impose specific restrictions.

Step S503: Test the trained initial quality inspection model based on the sample test data, and obtain model test parameters.

In some embodiments, the obtained sample test data is input into the trained initial quality inspection model to perform testing of the trained initial quality inspection model. After the test is completed, the model test of the initial quality inspection model can be obtained. parameter.

In some embodiments, the model test parameters are used to compare with a preset model compliance value. The model compliance value can be a range. If the obtained model test parameter value does not fall within this range, it represents the initial quality inspection model. If the test fails, you need to re-obtain sample training data and sample test data to update the model.

As shown in Figure 7, Figure 7 is an optional workflow configuration diagram of the quality inspection model management method provided by the embodiment of the present application. In order to improve the update efficiency of the quality inspection model, the relevant platform equipped with MLOps can be Relevant update steps are logically connected and form a visual workflow step. Furthermore, the platform includes a management end and a user end. The management end is responsible for configuring the workflow, the user end is responsible for displaying and running the workflow, and the management end includes a management interface. , the client includes the user interface.

For example, the left area in Figure 7 is the visualization area, and the right area is the configuration area. The workflow in the visualization area can include multiple sub-workflows, and each sub-workflow includes the following steps: Difficult case screening - Difficult case labeling -Data set-training evaluation-testing. The key step of each sub-workflow is the selection of difficult example data. Therefore, the configuration area is used to screen and label difficult example data for each sub-workflow. Specifically, the visualization table in the configuration area is used to select a specific visualization display method; the filtering range is used to represent the filtering range of difficult case data; the model name is used to represent the selection of the initial quality inspection model that needs to be updated and iterated; the annotation type is used to select The type of label required; the label template is used to adjust the label style of the visualization area. Among them, there is a judgment mark after each workflow node in the visualization area. If the node configuration passes, the judgment mark displays the configuration success status. If the node configuration fails, the judgment mark displays the configuration failure status. It should be noted that the workflow configured on the management side requires the user to set specific parameters before it can be executed.

As shown in Figure 8, Figure 8 is another optional workflow configuration diagram of the quality inspection model management method provided by the embodiment of the present application. After the workflow configuration of the quality inspection model is completed on the management end of the relevant platform equipped with MLOps , the corresponding workflow configuration of quality inspection model management can be displayed on the user interface, in which the training evaluation and testing of the model are displayed in the "Model Evaluation" area in Figure 8. Through the pre-configured workflow logic on the management side, relevant operators only need to click the component buttons corresponding to each step on the user side to complete the automated update operation of the quality inspection model. When the quality inspection accuracy of the initial quality inspection model decreases, There is no need to manually obtain relevant data and input it one by one into the initial quality inspection model to be updated, which reduces the professionalism of the operation and improves the update efficiency of the quality inspection model. In this way, the operator can immediately update the quality inspection model and get the update after The final quality inspection model is directly put into use, thereby improving the efficiency of related quality inspection work.

As shown in Figure 9, Figure 9 is an optional workflow execution diagram for quality inspection model management provided by the embodiment of this application. Since the configuration of the management end and the operation and display of the user end are all controlled by the workflow-server The back-end service is provided. Therefore, after configuring the workflow on the platform management side, you can configure specific parameters on the user side (user-web), and then perform specific operation steps through the back-end service of workflow-server. Details The process is as follows:

1. Configure the workflow in the management interface of the platform management side. The configuration work includes creating MLOps workflow templates, configuring virtual data sets and inference services (that is, configuring sample training sets and training methods of sample training sets), instantiating MLOps work, and Generate workflow configuration;

2. Start an update iteration on the client-web and run a certain step of the workflow, including obtaining the MLOps workflow configuration configured on the platform management side, displaying the instantiated MLOps steps according to the workflow configuration, and running the MLOps data set;

3. According to the operation of the client-web, find the corresponding trigger (Trigger) in the workflow-server back-end service and trigger the corresponding Component logic;

4. Component will call the corresponding service to complete the relevant logic. After completion, send the event to MessageQueue (MQ);

5. The workflow-controller obtains events from MQ and stores the execution results of the current step. The execution results can be persisted and the workflow can be advanced to the next workflow step.

Among them, Trigger is the execution service initiated by Trigger when the user clicks on a certain workflow step component; Component is the service that specifically executes a certain step, such as difficult case screening, training, evaluation, etc.; Controller is used to execute the flow of workflow status. , when a step is completed, it is responsible for storing the output of this step, reversing the status of the current step, and advancing to the next step.

In addition, in order to understand the above-mentioned embodiments of the present application more clearly, further examples will be provided below.

As shown in Figure 10, Figure 10 is an optional flow diagram of the quality inspection model management method provided by the embodiment of the present application. The following is a specific step description of this example:

Step 1, service deployment, first deploy the initial quality inspection model to the service scenario. The initial quality inspection model will reason about the initial quality inspection parameter data and generate return data;

Step 2, data analysis. From the returned data, statistics related model indicators can be used to determine whether the model effect has declined. If it has declined, model iteration is required. Specifically, model indicator statistics can be used to determine whether the model effect has declined. For the field of industrial quality inspection, by collecting production line data in real time, the accuracy of the model, defect detection rate, false point removal rate, through rate, and other indicators can be statistically calculated, and whether the indicators have declined can be judged through visualization;

Step 3. Model iteration is divided into the following steps: difficult case selection, data annotation, training, evaluation, A/B testing, online indicator observation, grayscale release, etc. Among them, model iteration can be run multiple times and the optimal one is selected. The target quality inspection model replaces the initial quality inspection model running online.

As shown in Figure 11, Figure 11 is another optional flow chart of the quality inspection model management method provided by the embodiment of the present application. The method in Figure 11 may include, but is not limited to, steps S601 to S605.

Step S601, display the quality inspection model management interface, and display the configuration components of the initial quality inspection model on the quality inspection model management interface, where the configuration components include a data configuration component, a model training component, a model testing component and a model publishing component;

In some embodiments, in order to further improve the efficiency of updating the quality inspection model, the update operation steps of the quality inspection model can be displayed in a visual manner.

As shown in Figure 9, the data configuration component includes a difficult case filtering component, a difficult case labeling component, and a data set component. Among them, click on the difficult case screening component to display the start time and completion time of the selected difficult case screening. Different start times and end times correspond to multiple different sets of difficult case data; click on the difficult case labeling component to display the selected difficult cases. Difficult example annotations corresponding to the data. Difficult example annotations can be annotated based on artificial intelligence models, or manually annotated by humans; click on the data set component to display one or more selected groups of difficult example data, and can also include general regular data.

In some embodiments, since the relevant sample data is preset and stored in the database, there is no need for operators to manually select and integrate it. At the same time, the relevant platform can perform automated testing based on the selected sample data, that is, a platform equipped with MLOps capabilities can perform Automatic data collection, automatic model monitoring, and through preset difficult case screening methods and preset model training and evaluation methods allow relevant operators to simply click on relevant components and select to start iterating components to automatically complete the initial quality inspection model The update greatly simplifies the operation steps of model update and improves the efficiency of quality inspection model update.

Step S602, in response to the click operation of the data configuration component, display the configured sample data on the quality inspection model management interface, where the sample data includes sample training data and sample test data;

In some embodiments, after clicking the data configuration component, the configured sample data can be displayed on the quality inspection model management interface. For example, sample training data and sample test data may be displayed in the area below the data set component of FIG. 8 .

Step S603, in response to the click operation of the model training component, the model training parameter values are displayed on the quality inspection model management interface. The model training parameter values are obtained after the initial quality inspection model is trained based on the sample training data;

In some embodiments, after clicking the model training component, the model training parameter values after the initial quality inspection model training can be displayed on the quality inspection model management interface. For example, the model training parameter values may include information such as model weights and loss functions.

Step S604: In response to the click operation of the model test component, the model test parameter values are displayed on the quality inspection model management interface. The model test parameter values are obtained after the initial quality inspection model is tested based on the sample test data, and the initial quality after passing the test is determined. The inspection model is the target quality inspection model;

In some embodiments, you can click the model testing component to perform testing of the initial quality inspection model, and display test-related model testing parameter values in the quality inspection model management interface. For example, the test parameter values may include model test values obtained after model testing, and preset model compliance values to facilitate intuitive data comparison with the model test values.

Step S605: In response to the click operation of the model release component, the released target quality inspection model is displayed on the quality inspection model management interface.

In some embodiments, after the initial quality inspection model passes the test, you can click the model release component to indicate that the target quality inspection model is officially launched. Alternatively, you can also set up a simulated online component. When the simulated online component is clicked, the online situation of the target quality inspection model can be simulated on some terminals to ensure the reliability and actual availability of the target quality inspection model before proceeding with the actual online launch. .

As shown in Figure 12, Figure 12 is an implementation flow chart of step S602 in Figure 11. In some embodiments, step S602 may include steps S701 to step S704:

Step S701, display the quality inspection model configuration interface, and display the data generation component and data annotation component of the sample data on the quality inspection model configuration interface;

In some embodiments, the quality inspection model configuration interface may be pop-up, that is, when sample data is not being obtained, it may be hidden in the quality inspection model management interface to enhance the simplicity of the interface.

In some embodiments, in order to configure the sample data, the quality inspection model configuration interface can be further displayed in the quality inspection model management interface, where the quality inspection model configuration interface includes a data generation component and a data annotation component, and the data generation component corresponds to the diagram The "Generate Difficult Examples" component in Figure 8 and the data annotation component correspond to the "Submit Annotation" component in Figure 8.

Step S702, in response to the click operation of the data generation component, the sample training data and sample test data obtained according to the initial quality inspection parameter configuration are displayed in the data configuration area of the quality inspection model configuration interface, where the sample training data includes difficult case data and regular data. ;

In some embodiments, click the data generation component to generate difficult example data, and display the configured sample training data and sample test data in the lower area of the "Dataset" component in Figure 8. Since the acquisition of difficult case data is more important in the quality inspection of target objects, the data generation component actually generates hard case data, and does not display the generation component of regular data, but is fixed in the database. Storage gets.

Step S703, in response to the click operation of the data annotation component, display the annotation results of the sample training data in the data annotation area of the quality inspection model configuration interface, where the annotation results include the difficult example data annotation results and the regular data annotation results;

In some embodiments, by clicking on the data annotation component, the annotation results of the sample training data can be displayed in the area below the "Submit annotation" component in Figure 8, where the annotation of the sample training data can be obtained through artificial intelligence annotation, or, Obtained by manual annotation by relevant staff.

Step S704, in response to the completion of the data configuration operation, display the configured sample data on the quality inspection model management interface according to the sample test data and the sample training data containing the annotation results.

In some embodiments, when sample training data, sample test data and corresponding annotation results are generated, the acquisition of sample data is completed.

As shown in Figure 13, Figure 13 is a schematic diagram of the functional modules of quality inspection model management provided by the embodiment of the present application. The embodiment of the present application also provides a quality inspection model management system that can implement the above quality inspection model management method, quality inspection model Management system includes:

The acquisition module 801 is used to obtain the initial quality inspection model, perform quality inspection operations on the target object according to the initial quality inspection model, and obtain initial quality inspection parameters;

The model training module 802 is used to filter and obtain sample training data from the database when it is monitored that the initial quality inspection parameters are not within the preset quality inspection parameter range, and train the initial quality inspection model based on the sample training data;

The model testing module 803 is used to filter and obtain sample test data from the database, test the trained initial quality inspection model based on the sample test data, and obtain model test parameters;

The target quality inspection model module 804 is used to determine the initial quality inspection model after testing as the target quality inspection model when the model test parameters match the preset model compliance value. The target quality inspection model is used to re-inspect the target object. operate.

The specific implementation of the quality inspection model management system is basically the same as the specific embodiment of the above-mentioned quality inspection model management method, and will not be described again here. On the premise of meeting the requirements of the embodiments of this application, the quality inspection model management system may also be provided with other functional modules to implement the quality inspection model management method in the above embodiments.

It should be noted that during the process of updating the initial quality inspection model, the decline in quality inspection accuracy of the initial quality inspection model can be automatically monitored by the platform, and the corresponding sample training data and sample test data are also pre-configured in the database. It does not require professionals to manually mine and filter data. After testing, the quality inspection model with the best test effect is selected as the target quality inspection model. It is understandable that even factory operators who are not very professional can complete the update of the quality inspection model independently, which improves the efficiency of the update of the quality inspection model. Moreover, the updated and determined target quality inspection model can be immediately applied to the quality inspection of the target object. This greatly improves the efficiency of the quality inspection work. If the problem of quality inspection accuracy decline occurs again in actual use, it can still be used. Update the quality inspection model again.

An embodiment of the present application also provides an electronic device. The electronic device includes a memory and a processor. The memory stores a computer program. When the processor executes the computer program, it implements the above quality inspection model management method. The electronic device can be any smart terminal including a tablet computer, a vehicle-mounted computer, etc.

As shown in Figure 14, Figure 14 is a schematic diagram of the hardware structure of an electronic device provided by an embodiment of the present application. The electronic device includes:

The processor 901 can be implemented by a general CPU (Central Processing Unit, central processing unit), a microprocessor, an application specific integrated circuit (Application Specific Integrated Circuit, ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement The technical solutions provided by the embodiments of this application;

The memory 902 can be implemented in the form of read-only memory (ReadOnlyMemory, ROM), static storage device, dynamic storage device, or random access memory (RandomAccessMemory, RAM). The memory 902 can store operating systems and other application programs. When implementing the technical solutions provided by the embodiments of this specification through software or firmware, the relevant program codes are stored in the memory 902 and called by the processor 901 to execute the implementation of this application. Example quality inspection model management method;

Input/output interface 903, used to implement information input and output;

Communication interface 904 is used to realize communication interaction between this device and other devices. Communication can be achieved through wired means (such as USB, network cable, etc.) or wirelessly (such as mobile network, WIFI, Bluetooth, etc.);

Bus 905, which transmits information between various components of the device (such as processor 901, memory 902, input/output interface 903, and communication interface 904);

The processor 901, the memory 902, the input/output interface 903 and the communication interface 904 implement communication connections between each other within the device through the bus 905.

Embodiments of the present application also provide a computer-readable storage medium. The computer-readable storage medium stores a computer program. When the computer program is executed by a processor, the above-mentioned quality inspection model management method is implemented.

As a non-transitory computer-readable storage medium, memory can be used to store non-transitory software programs and non-transitory computer executable programs. In addition, the memory may include high-speed random access memory and may also include non-transitory memory, such as at least one magnetic disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, the memory may optionally include memory located remotely from the processor, and the remote memory may be connected to the processor via a network. Examples of the above-mentioned networks include but are not limited to the Internet, intranets, local area networks, mobile communication networks and combinations thereof.

The embodiments described in the embodiments of the present application are for the purpose of more clearly illustrating the technical solutions of the embodiments of the present application, and do not constitute a limitation on the technical solutions provided by the embodiments of the present application. Those skilled in the art will know that with the evolution of technology and new technologies, As application scenarios arise, the technical solutions provided by the embodiments of this application are also applicable to similar technical problems.

Those skilled in the art can understand that the technical solutions shown in the figures do not limit the embodiments of the present application, and may include more or fewer steps than those shown in the figures, or combine certain steps, or different steps.

The device embodiments described above are only illustrative, and the units described as separate components may or may not be physically separate, that is, they may be located in one place, or they may be distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

Those of ordinary skill in the art can understand that all or some steps, systems, and functional modules/units in the devices disclosed above can be implemented as software, firmware, hardware, and appropriate combinations thereof.

The terms "first", "second", "third", "fourth", etc. (if present) in the description of this application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe specific objects. Sequence or sequence. It is to be understood that the data so used are interchangeable under appropriate circumstances so that the embodiments of the application described herein can be practiced in sequences other than those illustrated or described herein. Furthermore, the terms "include" and "having" and any variations thereof are intended to cover non-exclusive inclusions, e.g., a process, method, system, product, or apparatus that encompasses a series of steps or units and need not be limited to those explicitly listed. Those steps or elements may instead include other steps or elements not expressly listed or inherent to the process, method, product or apparatus.

It should be understood that in this application, "at least one (item)" and "several" refer to one or more, and "plurality" refers to two or more. "And/or" is used to describe the relationship between associated objects, indicating that there can be three relationships. For example, "A and/or B" can mean: only A exists, only B exists, and A and B exist simultaneously. , where A and B can be singular or plural. The character "/" generally indicates that the related objects are in an "or" relationship. “At least one of the following” or similar expressions thereof refers to any combination of these items, including any combination of a single item (items) or a plurality of items (items). For example, at least one of a, b or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c" ”, where a, b, c can be single or multiple.

In the several embodiments provided in this application, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are only illustrative. For example, the division of the above units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components may be combined or may be Integrated into another system, or some features can be ignored, or not implemented. On the other hand, the coupling or direct coupling or communication connection between each other shown or discussed may be through some interfaces, and the indirect coupling or communication connection of the devices or units may be in electrical, mechanical or other forms.

The units described above as separate components may or may not be physically separated. The components shown as units may or may not be physical units, that is, they may be located in one place, or they may be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

In addition, each functional unit in each embodiment of the present application can be integrated into one processing unit, each unit can exist physically alone, or two or more units can be integrated into one unit. The above integrated units can be implemented in the form of hardware or software functional units.

Integrated units may be stored in a computer-readable storage medium if they are implemented in the form of software functional units and sold or used as independent products. Based on this understanding, the technical solution of the present application is essentially or contributes to the existing technology, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium , including multiple instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods of various embodiments of the present application. The aforementioned storage media include: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), magnetic disk or optical disk, etc. that can store programs. medium.

The preferred embodiments of the embodiments of the present application have been described above with reference to the accompanying drawings, but this does not limit the scope of rights of the embodiments of the present application. Any modifications, equivalent substitutions and improvements made by those skilled in the art without departing from the scope and essence of the embodiments of the present application shall be within the scope of rights of the embodiments of the present application.

Claims (10)

1. A quality control model management method, the method comprising:

acquiring an initial quality inspection model, and performing quality inspection operation on a target object according to the initial quality inspection model to obtain initial quality inspection parameters;

when the initial quality inspection parameters are monitored to be not in the preset quality inspection parameter range, screening from a database and obtaining sample training data, and training the initial quality inspection model according to the sample training data;

screening and obtaining sample test data from a database, testing the initial quality inspection model after training according to the sample test data, and obtaining model test parameters;

when the model test parameters are matched with preset model standard reaching values, determining the initial quality inspection model after the test as a target quality inspection model, wherein the target quality inspection model is used for carrying out quality inspection operation on the target object again.

2. The quality control model management method according to claim 1, further comprising, after the initial quality control model after training is tested according to the sample test data and model test parameters are obtained:

when the model test parameters are not matched with the preset model reaching values, screening from a database again and obtaining sample training data and sample test data;

And carrying out updating training and updating test on the initial quality inspection model again according to the re-determined sample training data and the sample testing data, obtaining updated model testing parameters, and determining the updated initial quality inspection model as a target quality inspection model when the model testing parameters are matched with a preset model standard reaching value.

3. The quality inspection model management method according to claim 2, wherein when the initial quality inspection parameters are monitored not to be within the preset quality inspection parameter range, screening and obtaining sample training data from a database includes:

analyzing the initial quality inspection parameters to obtain a plurality of parameter categories of the initial quality inspection parameters, wherein each parameter category comprises a corresponding confidence coefficient, and the confidence coefficient is used for representing the credibility of the initial quality inspection model belonging to the parameter category;

and when the confidence coefficient is not in the preset quality inspection parameter range, determining a screening category of sample training data screened from the database according to the confidence coefficient, and selecting corresponding sample training data from the database according to the screening category.

4. A quality control model management method according to claim 3, wherein the sample training data includes refractory case data and regular data;

The training the initial quality inspection model according to the sample training data comprises the following steps:

activating a training trigger of the initial quality inspection model when the sample training data is determined from a database;

according to the training trigger in the activated state, the difficult-case data are input into the initial quality inspection model to obtain a difficult-case training result, and the conventional data are input into the initial quality inspection model to obtain a conventional training result;

and according to the difficult training result and the conventional training result, training the initial quality inspection model is completed.

5. The quality control model management method according to claim 4, wherein the screening and obtaining sample test data from a database, testing the trained initial quality control model according to the sample test data, and obtaining model test parameters includes:

activating a test trigger of the initial quality inspection model when training of the initial quality inspection model is completed;

screening and obtaining sample test data from a database according to the test trigger in an activated state;

and testing the initial quality inspection model after training according to the sample test data, and obtaining model test parameters.

6. The quality control model management method according to claim 1, wherein after the determining that the initial quality control model after training is a target quality control model, further comprising:

displaying a quality inspection model management interface, and displaying a configuration component of the initial quality inspection model on the quality inspection model management interface, wherein the configuration component comprises a data configuration component, a model training component, a model testing component and a model publishing component;

responding to clicking operation of a data configuration component, and displaying sample data obtained by configuration on the quality inspection model management interface, wherein the sample data comprises sample training data and sample test data;

responding to clicking operation of a model training component, and displaying model training parameter values on the quality inspection model management interface, wherein the model training parameter values are obtained after the initial quality inspection model is trained according to the sample training data;

responding to clicking operation of a model testing component, displaying model testing parameter values on the quality testing model management interface, wherein the model testing parameter values are obtained after the initial quality testing model is tested according to the sample testing data, and determining that the initial quality testing model after the test is passed is a target quality testing model;

And responding to clicking operation of the model release component, and displaying the released target quality inspection model on a quality inspection model management interface.

7. The quality control model management method according to claim 6, wherein displaying the configured sample data on the quality control model management interface includes:

displaying a quality inspection model configuration interface, and displaying a data generation component and a data annotation component of the sample data on the quality inspection model configuration interface;

responding to clicking operation of a data generating component, and displaying sample training data and sample test data obtained according to the initial quality inspection parameter configuration in a data configuration area of the quality inspection model configuration interface, wherein the sample training data comprises difficult case data and conventional data;

responding to clicking operation of a data labeling component, and displaying labeling results of the sample training data in a data labeling area of the quality inspection model configuration interface, wherein the labeling results comprise difficult-case data labeling results and conventional data labeling results;

and responding to the data configuration completion operation, and displaying the configured sample data on the quality inspection model management interface according to the sample test data and the sample training data containing the labeling result.

8. A quality control model management system, the system comprising:

the acquisition module is used for acquiring an initial quality inspection model, and performing quality inspection operation on a target object according to the initial quality inspection model to obtain initial quality inspection parameters;

the model training module is used for screening and obtaining sample training data from a database when the initial quality inspection parameters are monitored to be not in the preset quality inspection parameter range, and training the initial quality inspection model according to the sample training data;

the model test module is used for screening and obtaining sample test data from a database, testing the initial quality inspection model after training according to the sample test data, and obtaining model test parameters;

and the target quality inspection model module is used for determining the initial quality inspection model after the test as a target quality inspection model when the model test parameters are matched with preset model standard reaching values, and the target quality inspection model is used for carrying out quality inspection operation on the target object again.

9. An electronic device comprising a memory storing a computer program and a processor implementing the quality inspection model management method of any one of claims 1 to 7 when the computer program is executed by the processor.

10. A computer readable storage medium storing a computer program, characterized in that the computer program, when executed by a processor, implements the quality inspection model management method of any one of claims 1 to 7.