Scene instance generation method, electronic equipment, storage medium and computer program product

Abstract

The invention provides a scene instance generation method, electronic equipment, a storage medium and a computer programmer.The method can comprise the steps that according to scene description information, an algorithm instance and a control instance related to a scene are determined; the algorithm instance and the control instance are created in advance; the execution sequence of the algorithm instance and the control instance is determined, the scene instance is obtained, and the execution sequence is used for indicating at least one of the input data, the source of the input data, the output data and the destination of the output data of the algorithm instance and the control instance. According to the embodiment of the application, the algorithm instance and the control instance related to the scene description information can be automatically matched only by acquiring the scene description information input by the user, and the algorithm instance and the control instance can be determined to complete creation of the scene instance. Therefore, a user can automatically and conveniently generate the scene instance without mastering development skills, and the efficiency of the scene instance is improved.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a method for generating a scene instance, an electronic device, a storage medium, and a computer program product. Background Art

[0002] Big data platforms for generating scenario instances can provide data platform support for big data mining, big data analysis, and digital operations across product lines. Scenario instances are generated through the big data platform to address diverse scenarios. However, in related technologies, big data platforms place high demands on developers, requiring them to possess a high level of professional expertise and programming knowledge. This makes them unsuitable for the cost-effective and rapid delivery requirements of scenario instance construction. Summary of the Invention

[0003] Embodiments of the present application provide a method for generating a scenario instance, an electronic device, a storage medium, and a computer program product.

[0004] In a first aspect, an embodiment of the present application provides a method for generating a scenario instance, which may include the following steps:

[0005] According to the scenario description information, determine the algorithm instance and control instance related to the scenario; the algorithm instance and control instance are created in advance;

[0006] Determine the execution order of the algorithm instance and the control instance to obtain the scenario instance, where the execution order is used to indicate at least one of the input data, the source of the input data, the output data, and the destination of the output data of the algorithm instance and the control instance.

[0007] In a second aspect, an embodiment of the present application provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory, wherein the processor implements any of the above-described methods when executing the computer program.

[0008] In a third aspect, an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, it implements any of the methods described above.

[0009] In a fourth aspect, an embodiment of the present application provides a computer program product, including a computer program / instruction, which implements any of the above methods when executed by a processor.

[0010] Compared with the prior art, this application has the following advantages:

[0011] According to the embodiments of the present application, during the process of generating a scenario instance, the present application only needs to obtain the scenario description information input by the user, that is, the scenario requirements, to automatically match the algorithm instance and control instance related to the scenario description information, and then determine the algorithm instance and control instance to complete the creation of the scenario instance. As a result, the scenario instance can be automatically and conveniently generated without the user having to master the scene development skills, thereby improving the efficiency of the scenario instance.

[0012] The above description is only an overview of the technical solution of this application. In order to more clearly understand the technical means of this application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of this application more obvious and easy to understand, the specific implementation methods of this application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In the accompanying drawings, unless otherwise specified, the same reference numerals throughout the multiple drawings represent the same or similar components or elements. These drawings are not necessarily drawn to scale. It should be understood that these drawings only depict some embodiments according to the present application and should not be regarded as limiting the scope of the present application.

[0014] Figure 1 This is a flowchart of a method for generating a scene instance according to an embodiment of the present application;

[0015] Figure 2 This is a schematic diagram of a scene description information interface according to an embodiment of the present application;

[0016] Figure 3 is a schematic diagram of an algorithm example of an embodiment of the present application;

[0017] Figure 4 This is a schematic diagram of a completed scenario example according to an embodiment of the present application;

[0018] Figure 5 This is one of the interactive interface diagrams of an embodiment of the present application;

[0019] Figure 6 This is the second interactive interface diagram of an embodiment of the present application;

[0020] Figure 7 This is a schematic diagram of a communication protocol format of an example of an embodiment of the present application;

[0021] Figure 8 This is the third interactive interface diagram of an embodiment of the present application;

[0022] Figure 9 This is a schematic diagram of the principle of a method for generating a scene instance according to an embodiment of the present application;

[0023] Figure 10is a structural block diagram of a device for generating a scene instance according to an embodiment of the present application; and

[0024] Figure 11 A block diagram of an electronic device used to implement an embodiment of the present application. DETAILED DESCRIPTION

[0025] Hereinafter, only certain exemplary embodiments are briefly described. As will be appreciated by those skilled in the art, the described embodiments may be modified in various ways without departing from the spirit or scope of the present application. Therefore, the drawings and description are to be regarded as illustrative in nature and not restrictive.

[0026] To facilitate understanding of the technical solutions of the embodiments of the present application, the following describes the related technologies of the embodiments of the present application. The following related technologies can be combined with the technical solutions of the embodiments of the present application as optional solutions, and all of them fall within the scope of protection of the embodiments of the present application.

[0027] The embodiment of the present application provides a method for generating a scene instance, such as Figure 1 The flowchart of the method for generating a scene instance according to an embodiment of the present application may include:

[0028] Step S101: determining the algorithm instance and control instance related to the scene according to the scene description information; the algorithm instance and control instance are created in advance;

[0029] Step S102: Determine the execution order of the algorithm instance and the control instance to obtain the scenario instance, where the execution order is used to indicate at least one of the input data, the source of the input data, the output data, and the destination of the output data of the algorithm instance and the control instance.

[0030] The execution subject of this application may be a big data platform that generates scenario instances, and the big data platform is aimed at users who have scenario construction needs. For example, the scenario may be a scenario for counting passenger flow, a scenario for identifying vehicles that violate traffic rules, a scenario for identifying the position of text in an image or video, etc. The specific scenario may be defined by the user according to needs. The scenario instance may correspond to the algorithm instance and control instance involved in implementing the above scenario. The algorithm instance may be a code set or code module that implements a certain function. For example, the algorithm instance includes a code set or code module that implements a face capture function, a code set or code module that implements a liveness recognition function, etc. The control instance can be used to indicate the flow direction and processing method of the algorithm instance processing data. For example, the gateway control instance can be used to indicate that the processing method of the algorithm instance that implements the face capture function is based on the Kafka service. For example, a data storage control instance can be used to indicate that the data output by the algorithm instance is stored in the Clickhouse database.

[0031] The user's needs for the scene can be displayed through the scene description information. Figure 2 As shown, Figure 2 The generated scenario instance is used as an example of an AI scenario instance. Therefore, the scenario description information entered by the user corresponds to the steps of AI scenario generation. The user can enter the scenario description information in the dialog box corresponding to the scenario keyword, and can also enter the keywords of the algorithm they want to use in the dialog box corresponding to the algorithm keyword. After detecting that the user input is complete, the user input information can be parsed.

[0032] For example, the scenario description information can be a natural language segment, carried in the form of text or audio, and received by the big data platform. For example, the user-provided description information could be "Store customer flow statistics, need to count the number of people entering and exiting the store," "Identify vehicles parked outside the stop line for a long time," or "Identify the location of text or images." These three scenario descriptions can generate three corresponding scenario instances.

[0033] For scene description information, natural language processing technology (Neuro-Linguistic Programming) can be used to complete recognition based on keyword recognition, semantic analysis, etc. to obtain recognition results. Based on the recognition results, algorithm instances and control instances related to the scene can be determined in the pre-built algorithm instance library and control instance library.

[0034] The algorithm instances in the algorithm instance library and the control instances in the control instance library are pre-built and support editing and updating. Figure 3 For example, 3 contains 4 pre-built algorithm instances. Choose one to explain. Take the algorithm instance "passenger flow statistics" as an example. After building the algorithm instance, add library tags, names, types, descriptions, inputs (data), and outputs (data) and other descriptive information to the algorithm instance. Among them, the algorithm can also be called a module, so Figure 3 The module description is used in the module type. For example, if the module type is "Docker image," it can be used to indicate that the module (algorithm) is a single algorithm that performs the functions described in the module description. If the module type is "SDK," it can be used to indicate that the module (algorithm) contains multiple submodules that jointly perform the functions described in the module description.

[0035] After identifying the algorithm instances relevant to the scenario, the scenario's execution framework can be determined based on the scenario description. This execution framework corresponds to the data flow and processing methods between instances. Correspondingly, control instances can be used to constrain the data flow and processing methods, which is the function of the control instance. Based on the execution framework, an appropriate control instance can be selected from the control instance library. Pre-built control instances must contain at least the control instance name, type, and corresponding data flow.

[0036] After determining the algorithm instances and control instances related to the scenario, the execution order can be determined to complete the construction of the scenario instance. For example, the execution order is used to indicate at least one of the input data, the source of the input data, the output data, and the destination of the output data of the algorithm instance and the control instance.

[0037] Combine Figure 4 As shown, taking the scenario of store customer flow statistics as an example, the determined algorithm instances include completing store entry / exit statistics based on the customer flow statistics algorithm, and completing store entry gender and age statistics based on the facial attribute recognition algorithm. Therefore, based on store customer flow statistics, it is necessary to complete data collection, parsing and field storage of collected data, data analysis, etc. The above-mentioned parsing and field storage of collected data, data analysis, etc. can all be implemented through control instances. For example, the message gateway control instance can be used to obtain the customer flow statistics calculation and / or the facial attribute recognition algorithm output, such as the video acquisition device identification, pedestrian identification, entry and exit direction, gender, store identification, etc. Further, if it is an http request, the message gateway control is an API gateway control. The ESL (Entity-Subtype-k) control instance is used to implement the parsing of the collected data type and field name. The Clickhouse control instance is used to implement the storage of data according to the parsing results. The data analysis can be an instance of a neural network model class algorithm, which is used to perform result analysis based on the data. For example, based on pedestrian identification, entry and exit directions, and gender, we can analyze data such as the number of people entering the store during different time periods, the proportion of customers by gender, and the proportion of customers by age group. Finally, a database or real-time HTTP is used to store the analysis results.

[0038] Since each algorithm instance and each control instance are configured with the source of data and the source of output data when they are created in advance, the data situation of each instance can be determined based on the data collection, parsing and field storage of the collected data, data analysis, etc. that need to be completed in the scenario. For example, still taking the scenario of store customer flow statistics as an example, the input data of the customer flow statistics calculation and / or facial attribute recognition algorithm can be the video stream data output by the video acquisition device. The input data of the message gateway control can be the output data of the customer flow statistics calculation and / or facial attribute recognition algorithm. The input data of the ESL control instance can be the output data of the message gateway control. The input data of the Clickhouse control instance can be the output data of the ESL control instance. After the Clickhouse control instance, a data analysis algorithm instance can be set to perform data analysis based on the content output by the Clickhouse control instance to obtain results such as the store name, time period, number of people entering the store, male percentage, female percentage, age group percentage, etc. The results can be stored in a Mysql database.

[0039] Through the above process, when generating a scenario instance, this application only needs to obtain the scenario description information input by the user, that is, the scenario requirements, to automatically match the algorithm instance and control instance related to the scenario description information, and can determine the algorithm instance and control instance to complete the creation of the scenario instance. As a result, the scenario instance can be automatically and conveniently generated without the user having to master development skills, which improves the efficiency of the scenario instance and reduces the difficulty of scenario instance development.

[0040] In one embodiment, the step S101 of determining the scene-related algorithm instance based on the scene description information includes:

[0041] Step S1011: determining the characteristics of the scene description information;

[0042] Step S1012: Acquire features of each algorithm instance in the algorithm instance set; the features of the algorithm instance are used to indicate at least one of the algorithm name, algorithm classification, algorithm label, algorithm description, number of algorithm calls, and algorithm accuracy of each algorithm instance;

[0043] Step S1013: Determine the algorithm instance related to the scene based on the matching degree between the features of the scene description information and the features of each algorithm instance.

[0044] When determining the algorithm instance, the scene description information needs to be parsed first. For example, NLP technology can be used to complete the recognition through keyword recognition, semantic analysis, etc. to obtain the parsing result of the scene description information, that is, Figure 2The text corresponding to the scene keywords and algorithm keywords in the image is parsed to obtain a parsing result. The parsing result can be a vector representation of the text. Furthermore, a feature extraction model can be used to extract features from the parsing result to obtain a feature representation of the scene description information. Alternatively, the scene description information can be directly input into a pre-trained text feature extraction model to obtain the features of the scene description information.

[0045] Correspondingly, since the algorithm instance is pre-created and contains descriptive information such as label, name, type, description, input data, and output data, for each algorithm instance, a corresponding feature representation can be generated for each descriptive information such as label, name, type, description, input data, and output data, or an overall feature representation can be generated for each descriptive information such as label, name, type, description, input data, and output data.

[0046] Thus, after determining the features of the scene description information, the features of each algorithm instance in the algorithm instance library can be obtained. The features of the scene description information are then matched against the features of each algorithm instance to determine the algorithm instance associated with the scene. During matching, the algorithm instance with the highest matching similarity can be selected, or multiple algorithm instances with matching similarities exceeding a specified similarity threshold can be selected. Matching can be implemented using a matching model trained using feature samples of the scene description information and feature samples of the algorithm instances. Based on the matching degree, the matching model can output at least one algorithm instance that matches the features of the received scene description information according to the matching degree. The matching degree can be expressed as a percentage or a level. For example, algorithm instances with a similarity greater than 85% can be considered very similar, while those with a similarity less than 60% can be considered dissimilar. In other words, upon receiving at least one of the features of the scene description information and the features of the algorithm instance, the matching model can match algorithm instances in the algorithm instance library based on the received content, thereby selecting the algorithm instance with the highest matching degree with the received content, or selecting multiple algorithm instances with matching degrees exceeding a preset matching threshold.

[0047] Selecting the algorithm instance with the highest similarity as the one relevant to the scene can improve consistency with the scene description. Selecting multiple algorithm instances whose matching results are similar to those above a specified similarity threshold can provide users with diverse choices and facilitate scene expansion.

[0048] In one embodiment, the method may further include:

[0049] Step S103: supplementing the scene description information according to the determined algorithm instance related to the scene to obtain a supplemented result;

[0050] Step S104: The supplemented result is displayed in a visual form in the scene description information display area.

[0051] After determining the algorithm instance relevant to the scenario, the scenario description information can be supplemented based on the algorithm instance's description information to obtain a supplementary result. For example, using an algorithm instance that counts the number of people entering and exiting a store based on a customer flow counting algorithm, the output information can include camera ID, trigger time, pedestrian ID, entry and exit direction, and store door number. Accordingly, the output of camera ID, trigger time, pedestrian ID, entry and exit direction, and store door number can serve as supplementary information for the scenario description.

[0052] For example, using a facial recognition algorithm to calculate the gender and age of people entering a store, the algorithm can output the camera ID, trigger time, store door number, facial recognition ID, facial gender, and facial age. These values ​​can then serve as supplementary information for the scene description.

[0053] Taking the example of an algorithm that locates the target text based on a text recognition algorithm, the top left vertex coordinates, width, and height of the target text rectangle can be output. Correspondingly, the top left vertex coordinates, width, and height of the target text rectangle can be used as supplementary results for scene description information.

[0054] On the one hand, the supplementary results can be displayed in a visual form in the scene description information display area, thereby showing the user the scene's execution logic or a brief description of the scene's functions, making it easier for the user to understand the relevant situation of the scene. On the other hand, the supplementary results can serve as a basis for scene expansion. For example, if only relying on the scene description information, the number of matched algorithm instances may be limited or may not fully match the user's expectations. In this case, the description information of the matched algorithm instances related to the scene is used as the supplementary results of the scene description information. The supplementary results can then be used to conduct a secondary query in the algorithm instance library to obtain a richer set of algorithm instances.

[0055] Ultimately, richer algorithm examples can serve as a reference for diversifying scenario examples, achieving the effect of learning from one example and applying it to other situations.

[0056] In one embodiment, determining the scene-related control instance according to the scene description information involved in step S101 may include:

[0057] Step S1014: Determine a scenario execution framework using at least one of the scenario description information and the supplemented result;

[0058] Step S1015: Determine a control instance using the scenario-related algorithm instance and the scenario execution framework.

[0059] Both the scenario description information and the supplementary results can be in natural language. Based on the analysis of this natural language, the scenario's execution framework can be determined. The execution framework is determined in a similar manner to that of an algorithm instance. Specifically, natural language processing techniques or pre-trained feature determination models can be used to first determine the natural language features of the scenario. Next, the natural language features of the scenario are matched with the features of a pre-stored execution framework to determine the scenario's execution framework.

[0060] Each scene execution framework will contain at least one control instance corresponding to the execution framework. In this way, the control instance related to the scene can be determined. In addition, when determining the control instance, the algorithm instance will be used as a basis for adjustment or constraint. For example, the algorithm instance related to the scene includes a first algorithm instance and a second algorithm instance. The output data of the first algorithm instance is in the first format, and the output data of the second algorithm instance is in the second format. Then, when selecting a control instance, it is necessary to meet the requirements of a control instance that is compatible with receiving data in the first format and data in the second format. Correspondingly, if the control instance that matches the scene execution framework can only receive the first data, or can only receive the second data, it will be replaced. Replaced with a control instance that is compatible with receiving the first data and the second data.

[0061] In one embodiment, when multiple algorithm instances or multiple control instances related to a scene are determined according to the scene description information, the method further includes the following steps:

[0062] Step S105: Displaying the preferred instance in the instance display area; the preferred instance is an algorithm instance selected from multiple algorithm instances, and a control instance selected from multiple control instances;

[0063] Step S106: displaying the candidate instances in the instance selection area; the candidate instances are instances that are not preferred algorithm instances among the multiple algorithm instances, and instances that are not preferred control instances among the multiple control instances;

[0064] Step S107: In response to the selection instruction of the candidate instance in the instance selection area, the candidate instance is converted into the preferred instance.

[0065] When the number of algorithm instances related to the scenario is determined to be multiple, and the number of control instances related to the scenario is determined to be multiple, it is necessary to first determine a preferred instance. The preferred instance is the algorithm instance selected from the multiple algorithm instances, and the control instance selected from the multiple control instances.

[0066] Combine Figure 5As shown, taking the scene description information as an example of identifying the location of text or images, when multiple algorithm instances and multiple control instances are determined based on the scene description information, the big data platform can select a preferred instance from the multiple algorithm instances and multiple control instances, and display the preferred instance in the instance display area.

[0067] The method of selecting the preferred instance can be determined based on the measurement results of multiple dimensions, such as the number of times multiple algorithm instances are finally selected, the accuracy of multiple algorithm instances, the degree of matching between multiple algorithm instances and scene description information, etc. For example, each dimension will have a corresponding score, and the preferred instance can be selected based on the scores of each dimension. Therefore, after determining the algorithm instance belonging to the preferred instance, the corresponding control instance can be selected based on the algorithm instance belonging to the preferred instance and the scene description information, thereby determining the control instance belonging to the preferred instance. Finally, the preferred instance is displayed in the instance display area. The instance display area can display the execution order and data direction between the priority instances, thereby providing the user with an intuitive visual window. For example, in Figure 5 and Figure 6 In the example, the lines between different instances represent standard input / output, meaning that the connected instances can be called directly. The arrows indicate the order of calls: from the API gateway to image template matching, and from the API gateway to OCR text recognition. In the code, the API gateway instance comes first, followed by the algorithm instance.

[0068] In addition, it also includes an instance selection area. The instance selection area can be another window parallel to the instance display area. In the instance selection area, alternative instances that are related to the scene description information but cannot be used as preferred instances can be displayed. When the user's selection instruction is detected, the instance replacement can be executed according to the selection instruction. That is, the original preferred instance can be replaced with the selected alternative instance according to the selection instruction. Exemplarily, the selection instruction can be detecting that the user has selected the icon of the alternative instance and the icon of the preferred instance, so that the replacement is completed according to the selection instruction; or it can be detecting that the user has selected the alternative instance and then dragging it, and the icon of the selected alternative instance is dragged to the instance display area, and the overlap time with the icon of the preferred instance exceeds the specified time, so that the replacement can be completed according to the selection instruction; or it can be detecting that the user has issued a selection instruction through voice or gesture, and then completing the replacement according to the analysis result of the selection instruction.

[0069] The above is explained using the replacement of instances as an example. In actual scenarios, when the determined algorithm instances related to the scenario are multiple algorithm instances, it can also include responding to the screening operation of multiple algorithm instances, so that before selecting the preferred instance, some algorithm instances are deleted first. For example, the descriptive information given by the user can be "store customer flow statistics, the number of people entering and leaving the store needs to be counted", and three algorithm instances related to the scenario are determined, namely, an algorithm instance for completing the statistics of the number of people entering / exiting the store based on the customer flow statistics algorithm, an algorithm instance for completing the gender and age statistics of people entering the store based on the face attribute recognition algorithm, and an algorithm instance for completing the statistics of the length of time pedestrians stay in the target area of ​​the store based on the pedestrian detection algorithm and the data analysis algorithm. Then the user can first perform a preliminary screening operation on multiple algorithm instances, and can retain all of them, or choose to delete some of the algorithm instances. Finally, the preferred instance will be determined based on the algorithm instances retained after the user's screening operation. Corresponding Figure 2 The process of determining the preferred instance can correspond to the steps of AI scene editing.

[0070] Therefore, by displaying preferred examples in the example display area and alternative examples in the example selection area, users can intuitively see which preferred examples are selected by the big data platform and which are the alternative examples for users to adjust the scenario examples. Therefore, the alternative examples can help users explore more scenario possibilities and provide users with the opportunity to create secondary scenarios.

[0071] In one embodiment, the step S106 of displaying the candidate instances in the instance selection area may include:

[0072] Step S1061: Determine the category of the candidate instance based on the description information of the candidate instance;

[0073] Step S1062: using the category as a visual menu bar for interacting with the user, and displaying the visual menu bar in the first sub-area of ​​the instance selection area;

[0074] Step S1063: In response to the selection operation on the visualization menu bar, the candidate instances corresponding to the category are displayed, and the candidate instances corresponding to the category are displayed in the second sub-area of ​​the instance selection area.

[0075] As mentioned above, when building the algorithm instance library and the control instance library, descriptive information such as the name and type of each algorithm instance and each control instance is configured accordingly. Therefore, categories can be constructed based on the descriptive information, and each algorithm instance and control instance can be categorized.

[0076] Combine Figure 5As shown in the figure, the right area is the instance display area, and the left area is the instance selection area. The instance selection area is further divided into an upper half and a lower half. The lower half corresponds to the first sub-area of ​​the instance selection area, and the upper half corresponds to the second sub-area of ​​the instance selection area. Figure 5 The lower half of the instance selection area is the "Common Algorithm Categories," which are used to indicate the category of algorithm instances. In other words, each category can be used as a visual menu bar for user interaction. Figure 5 The upper half of the instance selection area is "Common Algorithms," which indicates algorithm instances within each category. For example, if you select "Face Recognition" under "Common Algorithm Category," the "Face Recognition" category might contain multiple common algorithm instances, such as face recognition, mask detection, beautification, and capture.

[0077] When it is detected that the user selects any category in the “common algorithm categories”, as a response to the selection operation of the visualization menu bar, multiple algorithm instances under the category can be displayed in the “common algorithms”.

[0078] Similarly, the control instances in the alternative instances can also be displayed in the same way. Figure 5 As shown, the embodiment of the present application also includes customized corresponding content. The so-called customization can be a scene example built and stored by the user.

[0079] That is to say, the big data platform can display different content according to the detected user selection instruction. For example, the user selection instruction can be a click instruction, a hover instruction, etc. Figure 5 In the case of "Algorithm Library", the classification of algorithm instances is displayed in the first sub-area. Furthermore, when it is detected that the user instruction selects the classification of algorithm instance as "Face Recognition", the second sub-area displays multiple commonly used algorithm instances under the "Face Recognition" category. Similarly, when it is detected that the user selects the instruction as Figure 6 In the case of "Control Instance Library", the commonly used control instance categories can be displayed in the first sub-area. Taking "WEB" as an example, if it is detected that the user selects "WEB", multiple commonly used control instances under the "WEB" category can be displayed in the second sub-area.

[0080] Through the above method, alternative instances can be presented in a visual interactive manner, making it convenient for users to query and select alternative instances.

[0081] In one embodiment, the method may further include:

[0082] Step S108: generating a custom scene in response to the confirmation operation of the preferred instance;

[0083] Step S109: Associating the scene description information of the customized scene with the customized scene.

[0084] As mentioned previously, the presence of alternative instances facilitates user adjustments to the preferred instance. When adjustments are made, an interactive window can be displayed to inquire whether the user has completed the adjustment. Upon receiving the user's confirmation of the adjusted preferred instance, a custom scenario can be generated. In other words, a custom scenario can be a scenario generated by the user independently adjusting the optimal instance recommended by the big data platform and saving it for later use.

[0085] After generating a custom scene, you also need to generate description information for the custom scene. For example, the description information generation process may include: obtaining the name, type, function description and other information of each instance in the scene instance, performing language processing such as keyword extraction and word order compilation based on the obtained information, and finally generating the scene description information of the custom scene. The description information of the scene may include the name of the scene, the functions that the scene can perform, etc. Associate the scene description information with the custom scene. Figure 2 The process of determining the custom scene can correspond to the steps of AI scene editing.

[0086] For example, for a scenario involving store visitor counting, the big data platform recommends an optimal instance that includes a facial recognition algorithm. Based on the actual scenario, the user simply replaces the facial recognition algorithm instance with a liveness detection algorithm instance. As a result, the user changes the optimal recommendation, and in response to the confirmation of the preferred instance, a custom scenario is generated. The scenario description for the custom scenario can then be adjusted to "completing store visitor counting based on liveness detection."

[0087] For another example, using the scenario of counting the number of people visiting a store, the big data platform recommends the optimal example, including a facial attribute recognition algorithm instance. It also matches the motor vehicle number detection algorithm instance and the vacant parking space detection algorithm instance. These two algorithm instances are selected as candidate examples. The big data platform categorizes the algorithm instances based on their categories. These categories are displayed in the first subarea of ​​the instance selection area. For example, if the vehicle number detection algorithm instance and the vacant parking space detection algorithm instance are classified as "face recognition," the big data platform associates the vehicle number detection algorithm instance and the vacant parking space detection algorithm instance with the "face recognition" category. If the user selects the "face recognition" category, the second subarea of ​​the instance selection area displays all matching algorithm instances that fall under this category. Based on this, the user selects the motor vehicle number detection algorithm instance and / or the vacant parking space detection algorithm instance, replacing the facial attribute recognition algorithm instance in the optimal example. The generated scenario instance is then changed to a "parking space detection instance scenario." This parking space detection scenario can now be used to detect vacant parking spaces in a parking lot. Correspondingly, the parking space detection instance scenario can also be used as a custom scenario. After generating a new custom scenario instance, text language processing technology can be used to generate a custom scenario instance description based on the descriptions of the algorithm instances and control instances within the custom scenario instance.

[0088] In one embodiment, the step S102 of determining the execution order of the algorithm instance and the control instance may include:

[0089] Determine the execution order of algorithm instances and control instances based on the scenario execution framework and communication protocol configuration information;

[0090] The scenario execution framework is determined according to the scenario description information, and the communication protocol configuration information includes the communication protocol configuration information of the algorithm instance and the communication protocol configuration information of the control instance.

[0091] Using natural language processing technology, we first determine the characteristics of the scene description letter. Then, we match the characteristics of the scene textual expression with the characteristics of the pre-stored execution framework to determine the scene execution framework.

[0092] According to the execution scenario framework, the control instance corresponding to the execution scenario framework will be matched. Each instance will contain communication protocol configuration information. For example, the communication protocol format of each instance can be as follows: Figure 6 The content shown includes the definition of field name, field type, field description and remarks, etc., and the order of each instance can be determined by parsing the communication protocol. Figure 2 The process of determining the custom scenario can correspond to the steps of generating the AI ​​solution.

[0093] In one embodiment, the method may further include:

[0094] The scenario instance is executed and deployed, where the execution and deployment includes generating a preview legend, generating an execution code, generating an application container engine image, and sending the generated application container engine to a scenario instance execution device.

[0095] After the scenario instance is generated, the process of deploying the scenario instance may also be included. Deployment may include running or storing.

[0096] Specifically, the deployment may include: in response to the user's preview keep option, generating a preview legend of the scenario instance. The preview legend may be in the format of an editable legend. In response to the user's code generation option, generating a call relationship source code containing the software package. It should be understood that if it is a third-party library interface, it also includes packaged library files. In response to the user's application container engine image option, each instance contained in the scenario instance is compiled and packaged to generate an application container engine (docker) image that can run independently. In response to the send option, the instance is deployed to the corresponding system platform as an independent service or docker image. In addition, it also includes responding to the result acquisition option, deploying the instance to the corresponding system platform as an independent service or docker image, and obtaining the results fed back by the system platform after the operation is completed.

[0097] For previewing legends, executing code or applying container engine images, it also supports downloading or generating HTML links. Figure 8 As shown, after the scenario instance is generated, the download and share interface will be displayed to the user. The download and share interface provides the user with content item options and operation item options. Among them, the content item options may include "solution legend", "solution copy" and "solution associated components / algorithm packages". Among them, the solution legend can correspond to Figure 5 and Figure 6 The content displayed in the example display area. The solution copy can Figure 5 and Figure 6 The content displayed in the instance display area is converted into text. The result of the conversion can include the name and function of each instance, as well as the overall process overview. The overall process overview can be generated based on the description information of each instance through text language processing technology. The solution-related components / algorithm packages can be Figure 5 and Figure 6The examples displayed in the example display area are converted into components or algorithms. Additionally, the action options can correspond to "Download" and "One-click Share." Downloading allows you to download the content item to your local computer, while One-click Share allows you to load the content item into an HTML link.

[0098] The following is an example to illustrate the process of generating a scene instance.

[0099] Get the user's scene description information.

[0100] For example, the scenario description information is: store customer flow statistics, which requires statistics on the number of people entering and leaving the store, the gender of people entering and leaving the store, and the age of people entering and leaving the store.

[0101] Complete user scenario descriptions based on existing algorithm libraries, control libraries, etc.

[0102] Exemplarily, the completed results include: completing the statistics of the number of people entering / exiting the store based on the customer flow statistics algorithm, and outputting the camera ID, trigger time, pedestrian ID, entry and exit direction, and store door number information; completing the gender and age statistics of people entering the store based on the face attribute recognition algorithm, and outputting the camera ID, trigger time, store door number information, face recognition ID, face gender, and face age value; completing the statistics of the length of time pedestrians stay in the target area of ​​the store based on the pedestrian detection algorithm and data analysis algorithm, and outputting the camera ID, trigger time, pedestrian ID, and length of stay.

[0103] Get the user's adjustment results.

[0104] For example, user adjustments include calculating the duration of pedestrian stays in a target area within a store based on pedestrian detection and data analysis algorithms. Consequently, the final algorithm examples include: Counting the number of people entering and exiting a store based on a customer flow counting algorithm, which outputs camera ID, trigger time, pedestrian ID, entry and exit direction, and door number; and Counting the gender and age of store entrants based on a facial recognition algorithm, which outputs camera ID, trigger time, door number, facial recognition ID, facial gender, and facial age.

[0105] The input of the algorithm for counting the number of people entering and leaving a store based on the customer flow counting algorithm is the video stream, and the output is the camera ID, trigger time, pedestrian ID, entry and exit direction, and store door number information.

[0106] The input of the algorithm instance for completing gender and age statistics of people entering the store based on the facial attribute recognition algorithm is the video stream, and the output is the camera ID, trigger time, store door number information, facial recognition ID, facial gender, and facial age value.

[0107] The above two algorithm instances are the first algorithms to be executed and do not include other predecessor instances.

[0108] Visual editing.

[0109] Based on the scenario description and supplementary results, the scenario execution framework is determined. Control instances are then determined based on the scenario execution framework. Control instances include message gateways, ESL, Clickhouse, data analysis, and databases / real-time HTTP. The message gateway includes multiple options, such as Kafka services, GRPC / HTTP services, and an HDFS file system for message storage. Based on the user's selection, a control instance is determined to execute the corresponding message gateway functionality.

[0110] The data transfer process for control instances includes sending customer flow statistics to the Kafka service. From the Kafka service to Clickhouse, this process includes ETL processing, primarily for message parsing and field storage. The data analysis module is used to output business data from Clickhouse. This function generates statistical results (including store name, time period, number of store visitors, percentage of males, percentage of females, and percentage of age groups) and maps data fields to corresponding business needs (for example, mapping store ID to store name). If multiple algorithm instances meet the execution requirements, priority rules can be used to determine which one is best. These priority rules can be preset or user-defined. The interfaces between instances should adhere to standardized rules to the greatest extent possible, including input content, input format, output content, and output format. Standardized and non-standardized instance interfaces are distinguished by identifiers to facilitate automatic system selection of similar modules. Non-standardized instances must be debugged and verified to be error-free before source code integration, deployment, and release can be performed with or without standardized tags.

[0111] The following is a sequence table of each algorithm instance and control instance in the scenario instance

[0112]

[0113]

[0114] At this point, the scene instance is generated.

[0115] In addition, controls are displayed on the interactive interface, allowing users to download all instances locally (on the big data platform) with a single click. After receiving the packaged results, the initialization engine program is launched locally, automatically opening a local initialization wizard interface to help users fill in all the configuration options required to initialize, deploy, and start the relevant modules. Examples include: The configuration page for the customer flow counting / face attribute module: For the customer flow counting algorithm module, configuration options include the video stream address, the IP address, port, account, password, and corresponding topic for the Kafka messaging gateway, and the target deployment location (IP address, path, account, and password); the configuration page for the ESL module: Configuration options include the IP address, port, account, password, and database name for Clickhouse, and the IP address, port, account, password, database name, and table name for MySQL. Once configuration is complete, the one-click package deployment begins. The system compiles and packages the modules on the designated machine and deploys them to the target location. If the modules in the architecture diagram are already independently runnable, such as Docker services, then only the package is needed. After all services have successfully started, the output of the data analysis module will be visible in the database.

[0116] Combine Figure 9 The generation process of the scene instance is still described by taking the scene instance as the IA scene instance.

[0117] Figure 9 The users in this section are those with AI needs. The big data platform receives scenario description information input by users through preset methods (such as mini-programs, apps, and web systems) and can provide users with code-level AI scenario solutions.

[0118] The big data platform is used to generate AI scenarios and solutions, and includes a scenario generation module, a recommendation module, a solution generation module, and a visualization component linkage module. The specific description is as follows:

[0119] Scenario generation module: Generates a complete description of the user scenario based on the scenario description information provided by the user; the complete description can be a more specific and richer scenario description.

[0120] Recommendation module: Extracts scene keywords based on scene description information and / or complete description. Uses keywords to match algorithms in the AI ​​algorithm library (matching keywords include algorithm name, algorithm classification, algorithm label, algorithm description information, number of algorithm calls, algorithm accuracy indicators, and other algorithm identifiers).

[0121] Solution generation module: Based on the scenario (scenario description information and / or complete description) and the matched algorithm, a solution with the algorithm name keyword is formed, and the visualization component linkage module is automatically associated to generate an initial linkage template.

[0122] Visual Component Linkage Module: Provides a visual editing window for solutions, allowing users to select algorithms, components, and intermediate steps in the solution. It also supports drag-and-drop interface linkage between algorithm instances, component instances, and custom scenarios.

[0123] AI Algorithm Library: A collection of algorithm instances that provides an application interface (API or SDK) for AI control instances. Each algorithm instance has an algorithm identifier, such as the algorithm name, algorithm category, algorithm tag, algorithm description, algorithm call count, and algorithm accuracy.

[0124] Component Library: A component library is a collection of control instances that provides control capabilities. It can include control instances for different fields and directions, such as mobile applications, industry cloud, and big data. Components correspond to the controls in the aforementioned embodiments. Each control instance has a corresponding control identifier, such as the control name, control category, control label, control description, and control call count.

[0125] The Custom Function Library is a collection of custom scenarios, providing an interface for custom scenarios. Custom scenarios are user-created, as opposed to templated scenarios. Each custom scenario is identified by a custom function, such as the scenario name, category, and description.

[0126] AI solution release and deployment: This section mainly includes three contents, as follows:

[0127] Solution code generation module: This module is responsible for connecting the algorithm instance interfaces, control instance interfaces, and custom scenario interfaces based on the final solution saved by the user to complete the preliminary construction of the solution logic code.

[0128] Solution deployment module: Generates code according to the received code generation instructions, completes resource allocation and scheduling, and completes code compilation, execution, and deployment.

[0129] Solution effect preview module: according to the received user instructions (such as deployment and release, effect preview), it controls the execution of real solution running effect preview, release or call sample effect display.

[0130] Image environment: Contains resources such as image packages, plug-ins required for code running, and verification data for algorithm instance interfaces.

[0131] Hardware facilities: including CPU, GPU, network, storage and other resources required for construction, deployment and release.

[0132] Corresponding to the application scenario and method of the method provided in the embodiment of the present application, the embodiment of the present application also provides a device for generating a scenario instance. Figure 10FIG. 1 is a block diagram of a device for generating a scene instance according to an embodiment of the present application. The device for generating a scene instance may include:

[0133] The sub-instance determination module 1001 is used to determine the algorithm instance and control instance related to the scene according to the scene description information; the algorithm instance and control instance are created in advance;

[0134] The scenario instance generation module 1002 is used to determine the execution order of the algorithm instance and the control instance to obtain the scenario instance. The execution order is used to indicate at least one of the input data, the source of the input data, the output data and the destination of the output data of the algorithm instance and the control instance respectively.

[0135] In one embodiment, the sub-instance determination module 1001 may include:

[0136] A scene feature determination submodule, used to determine the features of the scene description information;

[0137] An instance feature determination submodule is used to obtain features of each algorithm instance in the algorithm instance set; the features of the algorithm instance are used to indicate at least one of the algorithm name, algorithm classification, algorithm label, algorithm description, number of algorithm calls, and algorithm accuracy of the algorithm instance;

[0138] The algorithm instance determination submodule is used to determine the algorithm instance related to the scene based on the matching degree between the characteristics of the scene description information and the characteristics of the algorithm instance.

[0139] In one embodiment, the method may further include:

[0140] An information supplementation module is used to supplement the scene description information according to the determined algorithm instance related to the scene to obtain a supplementary result;

[0141] The information display module is used to display the supplementary results in a visual form in the scene description information display area.

[0142] In one embodiment, the sub-instance determination module 1001 may further include:

[0143] A scenario execution framework determination submodule, configured to determine a scenario execution framework using at least one of the scenario description information and the supplementary result;

[0144] The control instance determination submodule is used to determine the control instance by using the algorithm instance related to the scenario and the scenario execution framework.

[0145] In one embodiment, when multiple algorithm instances or multiple control instances related to a scene are determined according to the scene description information, an instance adjustment module is further included. The instance adjustment module may include:

[0146] The preferred instance display submodule is used to display the preferred instance in the instance display area; the preferred instance is an algorithm instance selected from multiple algorithm instances, and a control instance selected from multiple control instances;

[0147] The alternative instance display submodule is used to display the alternative instances in the instance selection area; the alternative instances are instances that are not preferred algorithm instances among multiple algorithm instances, and instances that are not preferred control instances among multiple control instances;

[0148] The adjustment execution submodule is used to respond to the selection instruction of the candidate instance in the instance selection area and convert the candidate instance into the preferred instance.

[0149] In one embodiment, the alternative example display submodule may include:

[0150] A category determination unit, configured to determine the category of the candidate instance based on the description information of the candidate instance;

[0151] A visual interaction unit, used to use categories as a visual menu bar for interacting with users;

[0152] The display execution unit is used to display the candidate instance corresponding to the category in response to a selection operation on the visual menu bar.

[0153] In one embodiment, a scene instance customization module is further included, which may include:

[0154] A custom scene generation submodule, configured to generate a custom scene in response to a confirmation operation of the preferred instance;

[0155] The information association submodule is used to associate the scene description information of the custom scene with the custom scene.

[0156] In one embodiment, the scenario instance generation module 1002 may be specifically configured to:

[0157] Determine the execution order of algorithm instances and control instances based on the scenario execution framework and communication protocol configuration information;

[0158] The scenario execution framework is determined according to the scenario description information, and the communication protocol configuration information includes the communication protocol configuration information of the algorithm instance and the communication protocol configuration information of the control instance.

[0159] In one embodiment, a post-processing module is further included, and the post-processing module is specifically configured to:

[0160] The scenario instance is executed and deployed, where the execution and deployment includes generating a preview legend, generating an execution code, generating an application container engine image, and sending the generated application container engine to a scenario instance execution device.

[0161] The functions of each module in each device of the embodiment of the present application can be found in the corresponding description of the above method, and have corresponding beneficial effects, which will not be repeated here. It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data need to comply with the relevant laws, regulations and standards of the relevant countries and regions, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0162] Figure 11 FIG. 1 is a block diagram of an electronic device for implementing an embodiment of the present application. Figure 11 As shown, the electronic device includes: a memory 1110 and a processor 1120. The memory 1110 stores a computer program that can be run on the processor 1120. When the processor 1120 executes the computer program, the method in the above embodiment is implemented. The number of the memory 1110 and the processor 1120 can be one or more.

[0163] The electronic device also includes:

[0164] The communication interface 1130 is used to communicate with external devices and perform data exchange transmission.

[0165] If the memory 1110, the processor 1120, and the communication interface 1130 are implemented independently, the memory 1110, the processor 1120, and the communication interface 1130 can be connected to each other via a bus and communicate with each other. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 11 Only one thick line is used in the diagram, but this does not mean that there is only one bus or one type of bus.

[0166] Optionally, in a specific implementation, if the memory 1110, the processor 1120 and the communication interface 1130 are integrated on a chip, the memory 1110, the processor 1120 and the communication interface 1130 can communicate with each other through an internal interface.

[0167] An embodiment of the present application provides a computer-readable storage medium storing a computer program, which implements the method provided in the embodiment of the present application when the program is executed by a processor.

[0168] An embodiment of the present application also provides a chip, which includes a processor for calling and executing instructions stored in the memory from the memory, so that a communication device equipped with the chip executes the method provided in the embodiment of the present application.

[0169] An embodiment of the present application also provides a chip, including: an input interface, an output interface, a processor and a memory. The input interface, the output interface, the processor and the memory are connected through an internal connection path. The processor is used to execute the code in the memory. When the code is executed, the processor is used to execute the method provided in the embodiment of the application.

[0170] It should be understood that the processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc. It is worth noting that the processor may be a processor that supports the Advanced RISC Machines (ARM) architecture.

[0171] Furthermore, optionally, the above-mentioned memory may include a read-only memory and a random access memory. The memory may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory may include a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may include a random access memory (RAM), which is used as an external cache. By way of example but not limitation, many forms of RAM are available. For example, static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDRSDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link DRAM (SLDRAM) and direct RAM bus random access memory (DR RAM).

[0172] In addition, an embodiment of the present invention provides a computer program product, including a computer program / instruction, which implements the method described in any of the above embodiments when executed by a processor.

[0173] In the above embodiments, all or part of the embodiments may be implemented using software, hardware, firmware, or any combination thereof. When implemented using software, all or part of the embodiments may be implemented in the form of a computer program product. A computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the processes or functions according to the present application are generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium or transferred from one computer-readable storage medium to another.

[0174] In the description of this specification, the reference terms "one embodiment," "some embodiments," "example," "specific example," or "some examples" mean that the specific features, structures, materials, or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. Moreover, the specific features, structures, materials, or characteristics described may be combined in any appropriate manner in any one or more embodiments or examples. In addition, those skilled in the art may combine and combine different embodiments or examples described in this specification, as well as features of different embodiments or examples, unless they are mutually inconsistent.

[0175] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features being referred to. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one such feature. Throughout the description of this application, "plurality" means two or more, unless otherwise specifically defined.

[0176] Any process or method described in the flowchart or otherwise described herein can be understood to represent a module, segment or portion of code comprising one or more executable instructions for implementing the steps of a specific logical function or process. The scope of the preferred embodiments of the present application includes other implementations in which the functions may be performed in a different order than shown or discussed, including performing the functions substantially simultaneously or in reverse order depending on the functions involved.

[0177] The logic and / or steps described in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, apparatus or device (such as a computer-based system, a system including a processor or other system that can fetch instructions from an instruction execution system, apparatus or device and execute instructions), or used in combination with such instruction execution systems, apparatuses or devices.

[0178] It should be understood that various parts of the present application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. All or part of the steps of the above embodiment method can be completed by instructing the relevant hardware through a program, which can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiment.

[0179] In addition, the functional units in the various embodiments of the present application may be integrated into a single processing module, or each unit may exist physically separately, or two or more units may be integrated into a single module. The aforementioned integrated modules may be implemented in the form of hardware or in the form of software functional modules. If the aforementioned integrated modules are implemented in the form of software functional modules and sold or used as independent products, they may also be stored in a computer-readable storage medium. The storage medium may be a read-only memory, a magnetic disk, or an optical disk, etc.

[0180] The above is merely an exemplary embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any person skilled in the art can easily conceive of various modifications or substitutions within the technical scope described in this application, and such modifications or substitutions should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. A method for generating a scene instance, characterized in that: include: Determining, based on the scenario description information, an algorithm instance and a control instance related to the scenario; the algorithm instance and the control instance are created in advance; Determine the execution order of the algorithm instance and the control instance to obtain a scenario instance, wherein the execution order is used to indicate at least one of the input data, the source of the input data, the output data, and the destination of the output data of the algorithm instance and the control instance respectively.

2. The method according to claim 1, characterized in that Determine, based on the scenario description information, an algorithm instance related to the scenario, including: determining characteristics of the scene description information; Obtaining features of each algorithm instance in the algorithm instance set; the features of the algorithm instance are used to indicate at least one of an algorithm name, an algorithm classification, an algorithm label, an algorithm description, a number of algorithm calls, and an algorithm accuracy rate of the algorithm instance; An algorithm instance related to the scene is determined based on a degree of matching between features of the scene description information and features of each of the algorithm instances.

3. The method according to claim 1 or 2, characterized in that Also includes: Supplementing the scene description information according to the determined algorithm instance related to the scene to obtain a supplemented result; The supplementary results are displayed in a visual form in the scene description information display area.

4. The method according to claim 3, characterized in that Determine, based on the scene description information, a control instance related to the scene, including: Determining a scenario execution framework using at least one of the scenario description information and the supplemented result; A control instance is determined using an algorithm instance related to the scenario and the scenario execution framework.

5. The method according to claim 1, wherein In the case of determining multiple algorithm instances or multiple control instances related to the scene according to the scene description information, the method further includes: Displaying a preferred example in an example display area; the preferred example is an algorithm example selected from the plurality of algorithm examples, and a control example selected from the plurality of control examples; Displaying alternative instances in the instance selection area; the alternative instances are instances that do not belong to the preferred algorithm instance among the multiple algorithm instances, and instances that do not belong to the preferred control instance among the multiple control instances; In response to a selection instruction for an alternative instance in the instance selection area, the alternative instance is converted into a preferred instance.

6. The method according to claim 5, characterized in that The displaying of the candidate instances in the instance selection area includes: Determining the category of the candidate instance based on the description information of the candidate instance; Using the category as a visual menu bar for interacting with the user, the visual menu bar is displayed in the first sub-area of ​​the instance selection area; In response to a selection operation on the visual menu bar, alternative instances corresponding to the category are displayed, and the alternative instances corresponding to the category are displayed in a second sub-area of ​​the instance selection area.

7. The method according to claim 5, characterized in that Also includes: In response to a confirmation operation of the preferred instance, generating a custom scenario; The scene description information of the custom scene is associated with the custom scene.

8. The method according to claim 1, characterized in that Determining the execution order of the algorithm instance and the control instance includes: Determining the execution order of the algorithm instance and the control instance according to the scenario execution framework and communication protocol configuration information; The scenario execution framework is determined according to the scenario description information, and the communication protocol configuration information includes the communication protocol configuration information of the algorithm instance and the communication protocol configuration information of the control instance.

9. The method according to claim 1, characterized in that Also includes: The scenario instance is executed and deployed, and the execution deployment includes generating a preview legend, generating an execution code, generating an application container engine image, and sending the generated application container engine to a scenario instance execution device.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory, wherein the processor implements the method according to any one of claims 1 to 9 when executing the computer program.

11. A computer-readable storage medium, wherein a computer program is stored in the computer-readable storage medium, and when the computer program is executed by a processor, the method according to any one of claims 1 to 9 is implemented.

12. A computer program product comprising a computer program / instructions, which, when executed by a processor, implement the steps of the method according to claims 1 to 9.

Images