Label task processing method and device

Abstract

The embodiments of the present specification provide a label task processing method and device, the method comprising receiving a label task processing instruction, obtaining a label task processing list based on the label task processing instruction; obtaining the label task to be executed from the label task processing list, and generating a label task execution form based on the label task to be executed; determining a corresponding label processing model based on the label task execution form, and processing the label task to be executed according to the label task execution form and the corresponding label processing model.

Description

Technical Field

[0001] The embodiments of this specification relate to the field of computer technology, and in particular to a tag task processing method. One or more embodiments of this specification also relate to a tag task processing apparatus, a computing device, and a computer-readable storage medium. Background Art

[0002] At present, in custom AI (English full name: Artificial Intelligence, Chinese full name: Artificial Intelligence, English abbreviation: AI) visual computing applications, there are often computing scenarios for secondary processing of labeled data, such as combining the primary labeled data generated by video structuring with the labeled data from other data sources for further processing and calculation to achieve scenarios such as same-person computing; therefore, with the increasing number of large-scale AI visual data processing scenarios, there is an urgent need to provide a universal labeling task processing method that can realize large-scale AI visual data processing, so as to support multiple data sources and flexible and changeable user-defined logic. Summary of the invention

[0003] In view of this, the present specification provides a tag task processing method. One or more embodiments of the present specification also relate to a tag task processing apparatus, a computing device, and a computer-readable storage medium to solve the technical defects existing in the prior art.

[0004] According to a first aspect of an embodiment of this specification, a tag task processing method is provided, including:

[0005] Receive a tag task processing instruction, and obtain a tag task processing list based on the tag task processing instruction;

[0006] Acquire the tag task to be executed from the tag task processing list, and generate a tag task execution list based on the tag task to be executed;

[0007] A corresponding label processing model is determined based on the label task execution list, and the label task to be executed is processed according to the label task execution list and the corresponding label processing model.

[0008] Optionally, generating a label task execution list based on the label task to be executed includes:

[0009] Determine the tag task execution statement corresponding to the tag task to be executed;

[0010] Determining parameter variables in the tag task execution statement based on the preset processing conditions of the tag task to be executed;

[0011] A label task execution list of the label task to be executed is generated based on the parameter variable.

[0012] Optionally, after generating a label task execution form based on the label task to be executed, the method further includes:

[0013] Determining the scheduling frequency of the tag task to be executed;

[0014] The next execution time of the tag task to be executed is generated based on the scheduling frequency, and the next execution time of the tag task to be executed is persisted.

[0015] Optionally, after persisting the next execution time of the tag task to be executed, the method further includes:

[0016] Based on the next execution time of the tag task to be executed, it is determined whether there is a tag task to be executed in the tag task processing list.

[0017] Optionally, after generating a label task execution form based on the label task to be executed, the method further includes:

[0018] Storing the tag task execution order in an execution order data pool;

[0019] Accordingly, determining the corresponding label processing model based on the label task execution order includes:

[0020] An execution instruction for the label task execution order is received, and when it is determined that there is a label task execution order to be executed in the execution order data pool, a corresponding label processing model is determined based on the label task execution order to be executed.

[0021] Optionally, after processing the to-be-executed label task according to the label task execution form and the corresponding label processing model, the method further includes:

[0022] The processing result of the tag task to be executed is persisted.

[0023] Optionally, before receiving the tag task processing instruction, the method further includes:

[0024] Receive a tag task generation instruction from a user, and determine a tag task creation interface based on the tag task generation instruction;

[0025] Returning the label task creation interface to the user, and receiving label task creation parameters input by the user based on the label task creation interface;

[0026] The label task is generated based on the label task creation parameters.

[0027] Optionally, before receiving the tag task processing instruction, the method further includes:

[0028] Receive a label task generation instruction sent by a user, wherein the label task generation instruction carries label task creation parameters;

[0029] The label task is generated based on the label task creation parameters.

[0030] Optionally, the receiving tag task processing instruction includes:

[0031] Receive tag task processing instructions according to preset time intervals.

[0032] Optionally, before obtaining the tag task to be executed and generating a tag task execution order based on the tag task to be executed, the method further includes:

[0033] A state modification instruction for a tag task in the tag task processing list is received, and the state of the tag task in the tag task list is adjusted based on the state modification instruction.

[0034] Optionally, the acquiring the to-be-executed label task from the label task processing list includes:

[0035] When it is determined that there is a label task to be executed according to the status of the label task in the label task processing list, the label task to be executed is acquired.

[0036] According to a second aspect of an embodiment of this specification, a label task processing device is provided, including:

[0037] A processing instruction receiving module is configured to receive a label task processing instruction and obtain a label task processing list based on the label task processing instruction;

[0038] An execution order generating module is configured to obtain the tag task to be executed from the tag task processing list, and generate a tag task execution order based on the tag task to be executed;

[0039] The task processing module is configured to determine the corresponding label processing model based on the label task execution list, and process the label task to be executed according to the label task execution list and the corresponding label processing model.

[0040] According to a third aspect of the embodiments of this specification, a method for generating visual labels based on artificial intelligence is provided, comprising:

[0041] receiving a visual label generation instruction, and generating a visual label generation task based on the visual label generation instruction;

[0042] receiving a label task execution instruction for the visual label generation task, and acquiring a label task processing list based on the label task execution instruction;

[0043] Acquire the visual label generation task from the label task processing list, and generate a label task execution order based on the visual label generation task;

[0044] Determine a corresponding artificial intelligence model based on the label task execution list, and process the visual label generation task according to the label task execution list and the artificial intelligence model to generate the visual label.

[0045] According to a fourth aspect of the embodiments of this specification, there is provided a visual label generation device based on artificial intelligence, including:

[0046] A generation instruction receiving module is configured to receive a visual label generation instruction and generate a visual label generation task based on the visual label generation instruction;

[0047] A task list acquisition module is configured to receive a label task execution instruction for the visual label generation task, and acquire a label task processing list based on the label task execution instruction;

[0048] An execution order generating module is configured to obtain the visual label generation task from the label task processing list, and generate a label task execution order based on the visual label generation task;

[0049] The visual label generation module is configured to determine the corresponding artificial intelligence model based on the label task execution list, and process the visual label generation task according to the label task execution list and the artificial intelligence model to generate the visual label.

[0050] According to a fifth aspect of an embodiment of this specification, a computing device is provided, including:

[0051] Memory and processor;

[0052] The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions, wherein the processor implements the steps of the label task processing method or the artificial intelligence-based visual label generation method when executing the computer-executable instructions.

[0053] According to the sixth aspect of the embodiments of this specification, a computer-readable storage medium is provided, which stores computer-executable instructions, which, when executed by a processor, implement the steps of the label task processing method or the artificial intelligence-based visual label generation method.

[0054] An embodiment of the present specification implements a label task processing method and device, the method comprising receiving a label task processing instruction, obtaining a label task processing list based on the label task processing instruction; obtaining the label task to be executed from the label task processing list, and generating a label task execution form based on the label task to be executed; determining a corresponding label processing model based on the label task execution form, and processing the label task to be executed according to the label task execution form and the corresponding label processing model; the label task processing method can support multiple data sources and flexible and changeable custom logic of label developers. In the specific implementation process, the label data processing logic customized by the label developer can be abstracted into a unified label processing model, so that the label developer can quickly and real-time generate custom AI visual labels based on multiple data sources and corresponding label processing models. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 This is an example diagram of a specific application scenario of a tag task processing method provided by an embodiment of this specification;

[0056] Figure 2 is a flowchart of a tag task processing method provided by an embodiment of this specification;

[0057] Figure 3 It is a schematic diagram of a label task execution unit of a label task processing method provided by an embodiment of this specification;

[0058] Figure 4 It is a process flow chart of label task processing when a label task processing method provided by an embodiment of this specification is applied to a custom visual label generation system;

[0059] Figure 5 It is a process flow chart of performing single processing on a label task when a label task processing method provided by an embodiment of the present specification is applied to a custom visual label generation system;

[0060] Figure 6 is a flow chart of a method for generating visual labels based on artificial intelligence provided by an embodiment of this specification;

[0061] Figure 7 It is a structural schematic diagram of a label task processing device provided by an embodiment of this specification;

[0062] Figure 8 It is a structural schematic diagram of a visual label generation device based on artificial intelligence provided by an embodiment of this specification;

[0063] Fig. 9 It is a structural block diagram of a computing device provided by an embodiment of this specification. DETAILED DESCRIPTION

[0064] Many specific details are described in the following description to facilitate a full understanding of this specification. However, this specification can be implemented in many other ways than those described herein, and those skilled in the art can make similar generalizations without violating the connotation of this specification, so this specification is not limited to the specific implementation disclosed below.

[0065] The terms used in one or more embodiments of this specification are only for the purpose of describing specific embodiments, and are not intended to limit one or more embodiments of this specification. The singular forms of "a", "said" and "the" used in one or more embodiments of this specification and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used in one or more embodiments of this specification refers to and includes any or all possible combinations of one or more associated listed items.

[0066] It should be understood that although the terms first, second, etc. may be used to describe various information in one or more embodiments of this specification, this information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of one or more embodiments of this specification, the first may also be referred to as the second, and similarly, the second may also be referred to as the first. Depending on the context, the word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".

[0067] First, the terms involved in one or more embodiments of this specification are explained.

[0068] Custom AI visual labels: generally refers to a new data element generated by a single structured data generated based on AI visual computing, which is then processed twice or multiple times in a specific application semantic scenario.

[0069] Customized visual label generation system: provides a fast label development platform that supports label developers to customize application logic and generation rules. Label developers can debug and publish their own application logic based on this platform, and can also view label task operation records.

[0070] SQL: English full name: Structured Query Language, Chinese full name: Structured Query Language, English abbreviation: SQL, is a database query and programming language.

[0071] API: English full name: Application Programming Interface, Chinese full name: Application Programming Interface, English abbreviation: API, is a set of pre-defined functions, or refers to the agreement for connecting different components of a software system.

[0072] In this specification, a label task processing method is provided. This specification also relates to a label task processing device, a computing device, and a computer-readable storage medium, which are described in detail one by one in the following embodiments.

[0073] See also Figure 1 , Figure 1 An example diagram showing a specific application scenario of a tag task processing method provided by an embodiment of this specification.

[0074] Figure 1 The application scenario includes a label developer 102 and a custom visual label generation system 104, and the label task processing method is applied in the custom visual label generation system 104.

[0075] Specifically, the label developer 102 generates a custom label task on the custom visual label generation system 104, wherein there may be multiple label developers 102;

[0076] The custom visual label generation system 104 adds the label task to the label task processing list, and when receiving the label task processing instruction, obtains the label task to be executed in the label task processing list, and then determines the basic data source corresponding to the label task based on the label task execution statement corresponding to the label task, and the label processing model determined by the label developer 102 of the label task for the label task, wherein the basic data source is obtained from the basic data source library provided by the custom visual label generation system 104 for all label developers 102. Specifically, the basic data source is based on video label data, such as videos of customers coming and going in stores, videos of vehicles and residents entering and leaving communities, and videos of employees and vehicles entering and leaving factories. Then the basic video label data of these videos can be labels of multiple video frames of these videos, that is, video labels that annotate the basic attributes of the gender, clothing, hairstyle of the characters in multiple video frames, and the color, model, license plate number, etc. of the vehicles; and the label processing model is the processing logic for the basic data source customized by the label developer 102, such as how to process and calculate the basic data source to achieve the label task;

[0077] The custom visual label generation system 104 executes the label tasks to be executed by the label developer 102 based on the determined basic data source and the label processing model to obtain the custom visual labels to be generated by the label developer 102. For example, the basic video label data in a video is calculated through the label processing model to count the custom visual labels such as the number of people without wearing masks in the video or the number of people entering and leaving the community on that day.

[0078] See also Figure 2 , Figure 2 A flowchart of a tag task processing method provided according to an embodiment of the present specification is shown, comprising the following steps:

[0079] Step 202: Receive a tag task processing instruction, and obtain a tag task processing list based on the tag task processing instruction.

[0080] Among them, the label task processing method is applied to a custom visual label generation system; therefore, in actual applications, the label task processing instruction can be initiated by the custom visual label generation system at a fixed time. For example, the custom visual label generation system is set to initiate a label task processing instruction every 10 minutes to implement a check and execution of the label tasks to be executed every 10 minutes, so as to avoid the label tasks to be executed not being processed in time, thereby improving the user experience of label developers using the label task processing method.

[0081] The specific implementation is as follows:

[0082] The receiving tag task processing instruction comprises:

[0083] Receive tag task processing instructions according to preset time intervals.

[0084] Specifically, if the label task processing instruction is initiated periodically by the custom visual label generation system, the label task processing method provided in the embodiment of this specification does not impose any limitation on the time interval for the custom visual label generation system to initiate the label task processing instruction. It can be set according to actual needs, and can be set to 2 minutes, 3 minutes or 10 minutes.

[0085] In the specific implementation, the label task processing instruction can also be initiated by the label developer. For example, after the label developer creates the label task in the custom visual label generation system, he clicks the execution control in the custom visual label generation system to initiate the label task processing instruction.

[0086] In actual applications, before receiving a label task processing instruction, a label task must be generated based on a custom visual label generation system. There are two ways to generate a label task. The first way to generate a label task is as follows:

[0087] Before receiving the tag task processing instruction, the method further includes:

[0088] Receive a tag task generation instruction from a user, and determine a tag task creation interface based on the tag task generation instruction;

[0089] Returning the label task creation interface to the user, and receiving label task creation parameters input by the user based on the label task creation interface;

[0090] The label task is generated based on the label task creation parameters.

[0091] Among them, the user can be understood as the label developer of the above-mentioned embodiment; and the label task generation instruction can be understood as the label task generation instruction triggered when the label developer clicks or performs other operations on the custom visual label generation system.

[0092] Specifically, after receiving the user's label task generation instruction, the label task creation interface is first determined based on the label task generation instruction, and then the label task creation interface is returned to the user, and the label task creation parameters filled in by the user in the label task creation interface are received. Finally, the label task is generated based on the label task creation parameters, wherein the label task creation parameters include but are not limited to label name, label description, label start time, label end time, applicable database, label generation calculation, and SQL (English full name: Structured Query Language, Chinese full name: Structured Query Language, English abbreviation: SQL) statement corresponding to the label task, etc.

[0093] In actual applications, after receiving the label task creation parameters entered by the user in the label task creation interface, click the generation control of the label task creation interface to generate the label task based on the input label task creation parameters. In order to improve the user experience, subsequent users can also modify the label task creation parameters in the generated label task.

[0094] Generating tag tasks through the tag task creation interface allows users to directly perceive their application logic and the generation of tag tasks visually, thereby enhancing user experience.

[0095] In addition, the second generation method for the labeling task is as follows:

[0096] Before receiving the tag task processing instruction, the method further includes:

[0097] Receive a label task generation instruction sent by a user, wherein the label task generation instruction carries label task creation parameters;

[0098] The label task is generated based on the label task creation parameters.

[0099] Specifically, it can be understood that the custom visual label generation system provides an API interface for users, receives label task generation instructions sent by users and carrying label task creation parameters through the API interface, and then directly generates label tasks in a quick way based on the label task creation parameters without the user's perception, thereby improving work efficiency.

[0100] In specific implementation, a tag task is generated, and after receiving a tag task processing instruction, a tag task processing list is obtained based on the tag task processing instruction, wherein the tag task processing list includes one or more tag tasks generated by tag developers according to tag task creation parameters.

[0101] Step 204: Acquire the tag task to be executed from the tag task processing list, and generate a tag task execution list based on the tag task to be executed.

[0102] In actual applications, the tag tasks to be executed can be understood as the tag tasks with an online status in the tag task processing list, the tag tasks with an offline or to-be-online status in the tag task processing list can be understood as the tag tasks that cannot be executed at present, and the tag tasks with an online status are the tag tasks whose execution time has been reached.

[0103] Specifically, when there are tag tasks to be executed in the tag task processing list, these tag tasks to be executed are obtained, and then a corresponding tag task execution order is generated based on each tag task to be executed.

[0104] In the specific implementation, in order to enable the label developer to customize the application logic and to debug their own application logic based on the custom visual label generation system, in the label task processing method, the label developer can adjust the status of the label task in its label task list to adapt to its own application logic and enhance the experience. The specific implementation method is as follows:

[0105] Before acquiring the tag task to be executed and generating a tag task execution list based on the tag task to be executed, the method further includes:

[0106] A state modification instruction for a tag task in the tag task processing list is received, and the state of the tag task in the tag task list is adjusted based on the state modification instruction.

[0107] Among them, the status modification instruction for the label task in the label task processing list can be issued by the label developer through a click or other trigger operation in the custom visual label generation system. After receiving the status modification instruction for the label task in the label task processing list, the custom visual label generation system can adjust the status of the label tasks in the label task list based on the status modification instruction, such as adjusting the status of the label task from offline to online, or adjusting the status of the label task from online to offline, which fully reflects the label developer's application logic for the variability of label tasks.

[0108] In combination with the above, after the tag developer adjusts the status of the tag task in the tag task processing list, the tag task to be executed can be judged based on the status of the tag task in the tag task processing list. The specific implementation method is as follows:

[0109] The acquiring the to-be-executed label task from the label task processing list comprises:

[0110] When it is determined that there is a label task to be executed according to the status of the label task in the label task processing list, the label task to be executed is acquired.

[0111] For a detailed description of the status of the tag task, please refer to the above embodiment, which will not be repeated here.

[0112] In addition, during specific implementation, generating a label task execution order based on the label task to be executed specifically includes:

[0113] Determine the tag task execution statement corresponding to the tag task to be executed;

[0114] Determining parameter variables in the tag task execution statement based on the preset processing conditions of the tag task to be executed;

[0115] A label task execution list of the label task to be executed is generated based on the parameter variable.

[0116] The tag task execution statement corresponding to the tag task to be executed is the tag task creation parameter when the tag task is generated; and the preset processing condition of the tag task to be executed can be understood as the scheduling frequency of each tag task to be executed when it is generated.

[0117] Specifically, based on the preset processing conditions of the label task to be executed, the parameter variables in the label task execution statement and the variable values ​​corresponding to the parameter variables can be determined; then the variable values ​​corresponding to the parameter variables replace the original parameter variables in the label task execution statement, and then based on the label task execution statement after the parameter variables are replaced, a label task execution order for the label task to be executed is generated.

[0118] For example, the name of the label task to be executed is: **store-*group-age, where **store is the application scenario of the label task, *group is the label developer, and age is the calculation object of the label task;

[0119] The tag task execution statement corresponding to the tag task to be executed is:

[0120] INSERT INTO TABLE_PERSON

[0121] SELECT

[0122] 'age'AS age,

[0123] '${fire-time}'::TIMESTAMPTZ AS gmt_update,

[0124] FROM AI_MODEL_PERSON

[0125] WHERE TRUE

[0126] AND gmt_modified>='${fire-time}'::TIMESTAMPTZ-INTERVAL'1hour'.

[0127] That is, the basic data source is obtained from "FROM AI_MODEL_PERSON", and the basic data source is calculated according to the label data processing application logic "AND gmt_modified>='${fire-time}'::TIMESTAMPTZ-INTERVAL'1hour'" customized by the label developer, and the calculation result is written to "TABLE_PERSON"; "AND gmt_modified>='${fire-time}'::TIMESTAMPTZ-INTERVAL'1hour'" means that the current time is calculated one hour forward as the last execution time of the label task. In specific implementation, the label of the current time and the last execution time of the label task is taken as the basic data source for calculation.

[0128] The preset processing conditions of the label task to be executed are: execute once every 10 minutes;

[0129] Then the parameter variable in the tag task execution statement determined based on the preset processing condition of the tag task to be executed is: fire-time;

[0130] Therefore, the actual value corresponding to the parameter variable is used to replace the parameter variable in the above label task execution statement. When the actual value corresponding to fire-time is "2020-05-06 11:20:05", the label task execution statement after the parameter variable is replaced is:

[0131] INSERT INTO TABLE_PERSON

[0132] SELECT

[0133] 'age'AS age,

[0134] '2020-05-06 11:20:05'::TIMESTAMPTZ AS gmt_update,

[0135] FROM AI_MODEL_PERSON

[0136] WHERE TRUE

[0137] AND gmt_modified>='2020-05-06 11:20:05'::TIMESTAMPTZ-INTERVAL'1hour'.

[0138] Then when the next tag task execution order is generated for the tag task to be executed, '2020-05-0611:20:05' will be replaced with '2020-05-06 11:30:05', and so on.

[0139] Finally, based on the tag task execution statement with the parameter variables replaced, a tag task execution list of the tag task to be executed is generated. Specifically, based on the tag task execution statement with the parameter variables replaced, a tag task execution list of the tag task to be executed is generated. Figure 3 , Figure 3 This is a schematic diagram of the interface for executing a single data pool. Figure 3 The tag task execution list including multiple tag task execution statements with replaced parameter variables is generated for the tag task to be executed.

[0140] In the embodiments of this specification, the label task execution statement is the label data processing logic customized by the label developer. When the label task is implemented, the label data processing logic customized by each label developer is replaced with parameter variables based on actual needs to generate a unified label task execution form. Subsequently, the custom visual label generation system can process the label task according to the label task execution form, realize the automation of label tasks, and improve the work efficiency of label developers.

[0141] In another embodiment of the present specification, after generating a label task execution form based on the label task to be executed, the method further includes:

[0142] Determining the scheduling frequency of the tag task to be executed;

[0143] The next execution time of the tag task to be executed is generated based on the scheduling frequency, and the next execution time of the tag task to be executed is persisted.

[0144] The scheduling frequency of the tag task to be executed is determined when the tag task is generated, for example, the tag task to be executed is executed every 30 minutes.

[0145] In actual applications, after generating a tag task execution order based on the tag task to be executed, the scheduling frequency of the tag task to be executed is determined, and then the next execution time of the tag task to be executed is generated according to the scheduling frequency, and the next execution time of the tag task to be executed is persisted; for example, the scheduling frequency of the tag task to be executed is once every 10 minutes, and the current execution time of the tag task to be executed is 12.20, then the next execution time of the tag task to be executed generated according to the scheduling frequency is 12.30, and then 12.30 is persisted, so that the tag task to be executed can be informed that it needs to be executed at 12.30 when it is executed next time. This persistence method can reduce the number of times database data is accessed, greatly increasing the execution speed of the system.

[0146] In the specific implementation, the custom visual label generation system can determine whether the label task should be online or offline based on the next execution time of the label task. The specific implementation method is as follows:

[0147] After persisting the next execution time of the tag task to be executed, the method further includes:

[0148] Based on the next execution time of the tag task to be executed, it is determined whether there is a tag task to be executed in the tag task processing list.

[0149] For example, if the next execution time of the label task is determined to be 12.30 and the current time is 12.20, then the status of the label task can be automatically adjusted to "to be online" in the label task list, indicating that the label task is an unexecutable label task; and if the next execution time of the label task is 12.30 and the current time is 12.30, then the status of the label task can be automatically adjusted to "online" in the label task list 1 minute in advance, indicating that the label task is a label task to be executed.

[0150] By timely adjusting the status of the tag tasks in the tag task processing list according to the next execution time of the tag tasks, it is possible to ensure that the tag tasks to be executed are retrieved and executed on time, and to realize automatic adjustment of the tag task status, thereby enhancing the user experience.

[0151] Step 206: Determine a corresponding label processing model based on the label task execution list, and process the label task to be executed according to the label task execution list and the corresponding label processing model.

[0152] Among them, the label processing model can be understood as the processing logic for label data customized by the label developer, that is, when processing the label task to be executed corresponding to the label task execution order, the label data processing logic customized by the label developer is applicable. For example, if the label task to be executed corresponding to the label task execution order is to count the number of people who do not wear masks on that day, then the label processing model is customized by the label developer, and secondary calculation of the basic video labels can be used to count the number of people who do not wear masks on that day.

[0153] In specific implementation, the label processing model can be connected to the custom visual label generation system as a standard interface and logic. After the corresponding label processing model is determined based on the label task execution order, the label task to be executed is processed according to the label task execution order and the corresponding label processing model. In actual applications, the label task execution order includes the execution time of the label task to be executed, the database corresponding to the basic data source, and the specific execution content of the label task to be executed. After determining the label processing logic of the label developer of the label task to be executed, the label task to be executed can be specifically executed based on the label processing logic and the label task execution order to generate the AI ​​visual label that the label developer ultimately needs to obtain based on the label task to be executed. Specifically, the basic data source can be unified by the custom visual label generation system It provides that the custom visual label generation system can uniformly store a relatively rich variety of basic data sources to provide them to the initiator of the label task, so as to increase the convenience of processing the label task, and the rich and diverse basic data sources can also support more label tasks, greatly enhancing the user experience. In actual use, the video label data based on the basic data source, for example, the basic data source includes basic labels of the community video: gender, hairstyle, clothing and other video labels, the label task to be executed is the number of takeaway people entering and leaving the community that day, and the label processing model is the processing logic for the video labels in the basic data source customized by the label developer. Then, based on the label processing logic and the label task execution unit, the label task to be executed can be executed to obtain the execution result, that is, the visual label of the number of takeaway people entering and leaving the community that day.

[0154] In the embodiments of this specification, the label task processing method can support multiple data sources and flexible custom logic of label developers. During the specific implementation process, the label data processing logic customized by the label developer can be abstracted into a unified label processing model, so that the label developer can quickly and real-time generate custom AI visual labels based on multiple data sources and corresponding label processing models.

[0155] In another embodiment of the present specification, after generating a label task execution form based on the label task to be executed, the method further includes:

[0156] Storing the tag task execution order in an execution order data pool;

[0157] Accordingly, determining the corresponding label processing model based on the label task execution order includes:

[0158] An execution instruction for the label task execution order is received, and when it is determined that there is a label task execution order to be executed in the execution order data pool, a corresponding label processing model is determined based on the label task execution order to be executed.

[0159] For executing a single data pool, see Figure 3 ,The execution order data pool stores multiple label task execution orders.

[0160] In actual applications, after a label task execution order is generated based on the label task to be executed, the label task execution order will be stored in an execution order data pool. When an execution instruction for the label task execution order is received and it is determined that there is a label task execution order to be executed in the execution order data pool, the label processing model corresponding to each label task execution order is determined based on the label task execution order to be executed, and the label tasks to be executed are processed in parallel according to the label task execution order and the corresponding label processing model.

[0161] Specifically, the execution instruction of the label task execution order can be generated by the custom visual label generation system at a scheduled time, or it can be generated by the custom visual label generation system triggered by the label developer. The specific setting depends on the actual application and is not limited here.

[0162] In the embodiments of this specification, a custom SQL execution form (i.e., a label task execution form) is generated by a label task on a regular basis as required. Each execution form logic is usually only run once. The custom part of the logic has been parameterized. Once generated, the custom visual label generation system will pick it up as soon as possible and execute it in parallel to achieve simple and fast processing of the label tasks to be executed.

[0163] In actual applications, in order to avoid the loss of the execution results of the label task to be executed when the custom visual label generation system fails, the execution results of the label task to be executed will be persisted to increase the security of the execution results of the label task to be executed. Subsequently, based on the execution results of the label task to be executed, other label tasks that have a blood dependency relationship with the label task to be executed can be executed, as described below:

[0164] After processing the to-be-executed label task according to the label task execution list and the corresponding label processing model, the method further includes:

[0165] The processing result of the tag task to be executed is persisted.

[0166] The following combination Figure 4 and Figure 5 The label task processing method provided in this specification is further described by applying it to a custom visual label generation system. Figure 4 A process flow chart of label task processing provided by an embodiment of the present specification is shown when a label task processing method is applied to a custom visual label generation system, specifically including the following steps:

[0167] Step 402: Start.

[0168] That is, the custom visual label generation system is started.

[0169] Step 404: Initialize system Etc / Cache / Lock, etc.

[0170] That is, the custom visual label generation system is initialized, such as connecting to the basic video label database, caching, memory checking, etc. Among them, Etc is etcetera, that is, file configuration is performed during initialization, etc. Cache is cache initialization, and Lock is lock-related contention task initialization to avoid task contention.

[0171] Step 406: Whether there are any valid tag tasks to be processed.

[0172] Specifically, after the custom visual label generation system is started and initialized, it will check whether there are label tasks to be executed in the label task processing list.

[0173] In actual applications, if yes, that is, there is a valid tag task to be processed, then step 408 is executed; if no, that is, there is no valid tag task to be processed, then step 410 is executed.

[0174] Step 408: Pick up the valid tag tasks, replace the SQL parameter variables as required, and generate the formal execution order to be executed into the execution order data pool.

[0175] Specifically, when there is a label task to be executed in the label task processing list, the label task to be executed is obtained, the label task execution statement corresponding to the label task to be executed is determined, the parameter variables in the label task execution statement are determined based on the preset processing conditions of the label task to be executed, and then a label task execution order of the label task to be executed is generated based on the parameter variables, and the label task execution order is stored in the execution order data pool.

[0176] Step 410: End.

[0177] That is, when there is no valid tag task to be processed, the processing flow of the tag task processing method is terminated.

[0178] Step 412: Generate the next execution time of the task according to the scheduling frequency requirement.

[0179] Specifically, after the tag task execution order is stored in the execution order data pool, the scheduling frequency of each tag task to be executed is determined, and then the next execution time of the corresponding tag task to be executed is generated based on the scheduling frequency of each tag task to be executed.

[0180] Step 414: The operation result is persisted.

[0181] Specifically, after the next execution time of the corresponding label task to be executed is generated based on the scheduling frequency of each label task to be executed, the next execution time of each label task to be executed is persisted.

[0182] The label task processing logic of the label task processing method provided in the embodiment of this specification includes five methods: label metadata description information, such as label task name, label scheduling information description, such as label task scheduling frequency, label task system parameter definition, such as label task system timed scheduling interval, label task lineage dependency, such as the lineage dependency between the video label obtained in this label task and other video labels, label task variability application logic definition, such as the label task can be online or offline at any time and other variability application logic, wherein the above five methods are all implemented by the label developer. Through the label task processing logic of the label task processing method, the label developer can conveniently debug and adjust the custom label generation plan, which is convenient, simple and fast to use.

[0183] It should be noted that the technical solution of the label task processing method and the technical solution of the above-mentioned label task processing method belong to the same concept. For the details not described in detail in the technical solution of the label task processing method, please refer to the description of the technical solution of the above-mentioned label task processing method.

[0184] See also Figure 5, Figure 5 A process flow chart of performing single processing on a label task when a label task processing method provided by an embodiment of the present specification is applied to a custom visual label generation system is shown, specifically including the following steps:

[0185] 502: Start.

[0186] That is, the custom visual label generation system is started.

[0187] Step 504: Initialize system Etc / Cache / Lock, etc.

[0188] That is, the custom visual label generation system is initialized, such as connecting to the basic video label database, caching, memory checking, etc. Among them, Etc is etcetera, that is, file configuration is performed during initialization, etc. Cache is cache initialization, and Lock is lock-related contention task initialization to avoid task contention.

[0189] Step 506: Whether there are any execution orders to be processed.

[0190] Specifically, after the custom visual label generation system is started and initialized, it will search the execution order data pool to see if there are any pending label task execution orders. If so, that is, there are pending execution orders, then execute step 508. If not, that is, there are no pending execution orders, then execute step 514.

[0191] Step 508: Pick up the execution order and throw it into the buffer concurrent thread pool.

[0192] Specifically, when there is a pending label task execution order in the execution order data pool, the custom visual label generation system will pick up the execution order and put the execution order into the buffer concurrent thread pool to wait for concurrent execution.

[0193] Step 510: Select a suitable data model (source) and request execution.

[0194] Specifically, a corresponding label processing model is selected for each execution order placed in the buffered concurrent thread pool waiting for concurrent execution, and then concurrent execution is requested for the execution orders in the buffered concurrent thread pool.

[0195] Step 512: Execute the result persistently.

[0196] Specifically, the execution results of the tag tasks to be executed corresponding to the execution orders concurrently executed in the buffered concurrent thread pool are persisted.

[0197] Step 514: End.

[0198] The processing logic of the label task execution sheet of the label task processing method provided in the embodiment of this specification is: the custom SQL execution sheet is generated by the label task on a regular basis as required, and each execution sheet logic is usually only run once; the custom part of the logic has been parameterized, and once generated, the custom visual label generation system will pick it up and execute it as soon as possible, and its interface definition is as follows: it can obtain label task related information, such as the start execution time of the label task, the real data source information, such as the basic video label data obtained from the database in the execution sheet, and the SQL logic after the parameterization of the current scheduling, such as the SQL logic after replacing the parameter variables and other interfaces; through the processing logic of the label task execution sheet of the label task processing method, the label developer can quickly and relatively real-time generate custom labels through the label task; and the label developer's custom data processing logic is abstracted into a unified label processing interface and model. The label task processing method can also support multiple data sources and flexible label developer custom logic, making it convenient, simple and fast for label developers to use.

[0199] It should be noted that the technical solution of the label task processing method belongs to the same concept as the technical solution of the above-mentioned label task processing method. For the details not described in detail in the technical solution of the label task processing method, please refer to the description of the technical solution of any of the above-mentioned label task processing methods.

[0200] See also Figure 6 , Figure 6 A flowchart of a method for generating visual labels based on artificial intelligence is shown in one embodiment of this specification, comprising the following steps:

[0201] Step 602: Receive a visual label generation instruction, and generate a visual label generation task based on the visual label generation instruction.

[0202] In specific implementation, visual labels can be understood as structured data generated once based on AI visual calculations, which is then processed twice or multiple times in combination with specific application scenarios to generate a new data element. The structured data generated once based on AI visual calculations is the basic label in the above-mentioned basic data source, that is, the basic video label; then one or more basic video labels are combined with specific application scenarios and processed twice or multiple times based on the artificial intelligence model to generate a new visual label.

[0203] For example, the basic video labels are the basic labels of vehicles entering and leaving the community surveillance video (for example, the color, size, and license plate number of the vehicle, etc.). The application scenario is to calculate the number of parking lots in the community on that day. At this time, the basic labels of vehicles entering and leaving the community surveillance video can be processed through the artificial intelligence model to generate new visual labels (the number of parking lots in the community on 2020-1-1 was 10).

[0204] Step 604: Receive a label task execution instruction for the visual label generation task, and obtain a label task processing list based on the label task execution instruction.

[0205] Step 606: Acquire the visual label generation task from the label task processing list, and generate a label task execution order based on the visual label generation task.

[0206] Step 608: Determine a corresponding artificial intelligence model based on the label task execution list, and process the visual label generation task according to the label task execution list and the artificial intelligence model to generate the visual label.

[0207] Specifically, the artificial intelligence model is any machine learning model that can perform secondary or multiple processing on basic video labels to generate visual labels according to the visual label generation task, and the artificial intelligence model can be flexibly selected according to the actual application scenario, and no limitation is made here.

[0208] The artificial intelligence-based visual label generation method provided in the embodiments of this specification is applied in AI visual computing. After receiving the visual label generation instruction, the corresponding visual label generation task is generated. Through the customized visual label generation system and artificial intelligence model, label developers can quickly and relatively real-time generate any customized AI label.

[0209] It should be noted that the technical solution for processing the visual label generation task in the artificial intelligence-based visual label generation method and the technical solution for processing the label task to be executed in the above-mentioned label task processing method belong to the same concept. For details not described in detail in the technical solution for visual label generation based on artificial intelligence, please refer to the description of the technical solution of any of the above-mentioned label task processing methods.

[0210] Corresponding to the above method embodiment, this specification also provides a label task processing device embodiment, Figure 7 A schematic diagram of the structure of a label task processing device provided by an embodiment of the present specification is shown.

[0211] like Figure 7 As shown, the device comprises:

[0212] The processing instruction receiving module 702 is configured to receive a label task processing instruction and obtain a label task processing list based on the label task processing instruction;

[0213] An execution order generating module 704 is configured to obtain the tag task to be executed from the tag task processing list, and generate a tag task execution order based on the tag task to be executed;

[0214] The task processing module 706 is configured to determine a corresponding label processing model based on the label task execution list, and process the label task to be executed according to the label task execution list and the corresponding label processing model.

[0215] Optionally, the execution order generating module 704 is further configured to:

[0216] Determine the tag task execution statement corresponding to the tag task to be executed;

[0217] Determining parameter variables in the tag task execution statement based on the preset processing conditions of the tag task to be executed;

[0218] A label task execution list of the label task to be executed is generated based on the parameter variable.

[0219] Optionally, the device further includes:

[0220] A frequency determination module, configured to determine the scheduling frequency of the tag task to be executed;

[0221] The execution time generation module is configured to generate the next execution time of the tag task to be executed based on the scheduling frequency, and persist the next execution time of the tag task to be executed.

[0222] Optionally, the device further includes:

[0223] The task determination module is configured to determine whether there is a tag task to be executed in the tag task processing list based on the next execution time of the tag task to be executed.

[0224] Optionally, the device further includes:

[0225] A storage module, configured to store the tag task execution order in an execution order data pool;

[0226] Accordingly, the task processing module 706 is further configured to:

[0227] An execution instruction for the label task execution order is received, and when it is determined that there is a label task execution order to be executed in the execution order data pool, a corresponding label processing model is determined based on the label task execution order to be executed.

[0228] Optionally, the device further includes:

[0229] The persistence module is configured to persist the processing result of the tag task to be executed.

[0230] Optionally, the device further includes:

[0231] An interface determination module is configured to receive a tag task generation instruction from a user and determine a tag task creation interface based on the tag task generation instruction;

[0232] a parameter receiving module, configured to return the label task creation interface to the user, and receive label task creation parameters input by the user based on the label task creation interface;

[0233] The first task generating module is configured to generate the label task based on the label task creating parameters.

[0234] Optionally, the device further includes:

[0235] A generation instruction receiving module is configured to receive a label task generation instruction sent by a user, wherein the label task generation instruction carries label task creation parameters;

[0236] The second task generating module is configured to generate the label task based on the label task creating parameters.

[0237] Optionally, the processing instruction receiving module 702 is further configured to:

[0238] Receive tag task processing instructions according to preset time intervals.

[0239] Optionally, the device further includes:

[0240] The modification instruction receiving module is configured to receive a status modification instruction for a tag task in the tag task processing list, and adjust the status of the tag task in the tag task list based on the status modification instruction.

[0241] Optionally, the execution order generating module 704 is further configured to:

[0242] When it is determined that there is a label task to be executed according to the status of the label task in the label task processing list, the label task to be executed is acquired.

[0243] The above is a schematic scheme of a label task processing device of this embodiment. It should be noted that the technical scheme of the label task processing device and the technical scheme of the label task processing method described above belong to the same concept, and the details not described in detail in the technical scheme of the label task processing device can be found in the description of the technical scheme of the label task processing method described above.

[0244] Corresponding to the above-mentioned embodiment of the method for generating visual labels based on artificial intelligence, this specification also provides an embodiment of a device for generating visual labels based on artificial intelligence, Figure 8 FIG. 1 shows a schematic diagram of a structure of a visual label generation device based on artificial intelligence provided by an embodiment of the present specification. Figure 8 As shown, the device comprises:

[0245] A generation instruction receiving module 802 is configured to receive a visual label generation instruction and generate a visual label generation task based on the visual label generation instruction;

[0246] The task list acquisition module 804 is configured to receive a label task execution instruction for the visual label generation task, and acquire a label task processing list based on the label task execution instruction;

[0247] An execution order generating module 806 is configured to obtain the visual label generation task from the label task processing list, and generate a label task execution order based on the visual label generation task;

[0248] The visual label generation module 808 is configured to determine the corresponding artificial intelligence model based on the label task execution list, and process the visual label generation task according to the label task execution list and the artificial intelligence model to generate the visual label.

[0249] The artificial intelligence-based visual label generation device provided in the embodiments of this specification is applied in AI visual computing. After receiving the visual label generation instruction, the corresponding visual label generation task is generated. Through the customized visual label generation system and artificial intelligence model, label developers can quickly and relatively real-time generate any customized AI label.

[0250] The above is a schematic scheme of a visual label generation device based on artificial intelligence in this embodiment. It should be noted that the technical scheme of the visual label generation device based on artificial intelligence and the technical scheme of the visual label generation method based on artificial intelligence belong to the same concept. The details not described in detail in the technical scheme of the visual label generation device based on artificial intelligence can be found in the description of the technical scheme of the visual label generation method based on artificial intelligence.

[0251] Fig. 9 The block diagram of a computing device 900 according to an embodiment of the present specification is shown. The components of the computing device 900 include but are not limited to a memory 910 and a processor 920. The processor 920 is connected to the memory 910 via a bus 930, and a database 950 is used to store data.

[0252] The computing device 900 also includes an access device 940 that enables the computing device 900 to communicate via one or more networks 960. Examples of these networks include a public switched telephone network (PSTN), a local area network (LAN), a wide area network (WAN), a personal area network (PAN), or a combination of communication networks such as the Internet. The access device 940 may include one or more of any type of network interface (e.g., a network interface card (NIC)) that is wired or wireless, such as an IEEE 802.11 wireless local area network (WLAN) wireless interface, a World Wide Interoperability for Microwave Access (Wi-MAX) interface, an Ethernet interface, a universal serial bus (USB) interface, a cellular network interface, a Bluetooth interface, a near field communication (NFC) interface, and the like.

[0253] In one embodiment of the present specification, the above components of the computing device 900 and Fig. 9 Other components not shown in the figure may also be connected to each other, for example, via a bus. It should be understood that Fig. 9 The computing device structure block diagram shown is only for the purpose of illustration, and is not intended to limit the scope of this specification. Those skilled in the art can add or replace other components as needed.

[0254] The computing device 900 may be any type of stationary or mobile computing device, including a mobile computer or mobile computing device (e.g., a tablet computer, a personal digital assistant, a laptop computer, a notebook computer, a netbook, etc.), a mobile phone (e.g., a smart phone), a wearable computing device (e.g., a smart watch, smart glasses, etc.), or other types of mobile devices, or a stationary computing device such as a desktop computer or PC. The computing device 900 may also be a mobile or stationary server.

[0255] Among them, the processor 920 is used to execute the following computer-executable instructions, wherein the processor implements the steps of the label task processing method or the artificial intelligence-based visual label generation method when executing the computer-executable instructions.

[0256] The above is a schematic scheme of a computing device of this embodiment. It should be noted that the technical scheme of the computing device and the technical scheme of the above-mentioned label task processing method or the visual label generation method based on artificial intelligence belong to the same concept, and the details not described in detail in the technical scheme of the computing device can be referred to the description of the technical scheme of the above-mentioned label task processing method or the visual label generation method based on artificial intelligence.

[0257] An embodiment of the present specification also provides a computer-readable storage medium storing computer instructions, which, when executed by a processor, implement the steps of the label task processing method or the artificial intelligence-based visual label generation method.

[0258] The above is a schematic scheme of a computer-readable storage medium of this embodiment. It should be noted that the technical scheme of the storage medium and the technical scheme of the above-mentioned label task processing method or the visual label generation method based on artificial intelligence belong to the same concept, and the details not described in detail in the technical scheme of the storage medium can be referred to the description of the technical scheme of the above-mentioned label task processing method or the visual label generation method based on artificial intelligence.

[0259] The above is a description of a specific embodiment of the specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0260] The computer instructions include computer program codes, which may be in source code form, object code form, executable files or some intermediate forms, etc. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium, etc. It should be noted that the content contained in the computer-readable medium may be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electric carrier signals and telecommunication signals.

[0261] It should be noted that, for the above-mentioned method embodiments, for the sake of simplicity of description, they are all expressed as a series of action combinations, but those skilled in the art should be aware that the embodiments of this specification are not limited by the order of the actions described, because according to the embodiments of this specification, some steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the embodiments of this specification.

[0262] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0263] The preferred embodiments of this specification disclosed above are only used to help explain this specification. The optional embodiments do not describe all the details in detail, nor do they limit the invention to only the specific implementation methods described. Obviously, many modifications and changes can be made according to the content of the embodiments of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the embodiments of this specification, so that technicians in the relevant technical field can well understand and use this specification. This specification is only limited by the claims and their full scope and equivalents.

Claims

1. A label task processing method, comprising: Receive a tag task processing instruction, and obtain a tag task processing list based on the tag task processing instruction; Obtain the tag task to be executed from the tag task processing list, and generate a tag task execution list based on the tag task to be executed, wherein the generation of the tag task execution list based on the tag task to be executed includes: determining the tag task execution statement corresponding to the tag task to be executed; determining the parameter variables in the tag task execution statement based on the preset processing conditions of the tag task to be executed; generating the tag task execution list of the tag task to be executed based on the parameter variables; wherein the preset processing condition is the scheduling frequency; the parameter variable is the execution time; Determine the corresponding label processing model based on the label task execution form, and process the label task to be executed according to the label task execution form and the corresponding label processing model, wherein the label processing model is a processing logic for label data customized by a label developer, and determine the corresponding label processing model based on the label task execution form, and process the label task to be executed according to the label task execution form and the corresponding label processing model, including: determine the label processing model called by the label task execution form, and process the label task to be executed according to the label task execution form and the corresponding label processing model.

2. The tag task processing method according to claim 1, after generating a tag task execution list based on the tag task to be executed, further comprising: Determining the scheduling frequency of the tag task to be executed; The next execution time of the tag task to be executed is generated based on the scheduling frequency, and the next execution time of the tag task to be executed is persisted.

3. The tag task processing method according to claim 2, after persisting the next execution time of the tag task to be executed, further comprising: Based on the next execution time of the tag task to be executed, it is determined whether there is a tag task to be executed in the tag task processing list.

4. The tag task processing method according to claim 1, after generating a tag task execution list based on the tag task to be executed, further comprising: Storing the tag task execution order in an execution order data pool; Accordingly, determining the corresponding label processing model based on the label task execution order includes: An execution instruction for the label task execution order is received, and when it is determined that there is a label task execution order to be executed in the execution order data pool, a corresponding label processing model is determined based on the label task execution order to be executed.

5. The label task processing method according to claim 4, after processing the label task to be executed according to the label task execution list and the corresponding label processing model, further comprising: The processing result of the tag task to be executed is persisted.

6. The tag task processing method according to claim 1, before receiving the tag task processing instruction, further comprising: Receive a tag task generation instruction from a user, and determine a tag task creation interface based on the tag task generation instruction; Returning the label task creation interface to the user, and receiving label task creation parameters input by the user based on the label task creation interface; The label task is generated based on the label task creation parameters.

7. The tag task processing method according to claim 1, before receiving the tag task processing instruction, further comprising: Receive a label task generation instruction sent by a user, wherein the label task generation instruction carries label task creation parameters; The label task is generated based on the label task creation parameters.

8. The tag task processing method according to claim 1, wherein receiving a tag task processing instruction comprises: Receive tag task processing instructions according to preset time intervals.

9. The tag task processing method according to claim 1, before obtaining the tag task to be executed and generating a tag task execution list based on the tag task to be executed, further comprising: A state modification instruction for a tag task in the tag task processing list is received, and the state of the tag task in the tag task processing list is adjusted based on the state modification instruction.

10. The label task processing method according to claim 9, wherein obtaining the label task to be executed from the label task processing list comprises: When it is determined that there is a label task to be executed according to the status of the label task in the label task processing list, the label task to be executed is acquired.

11. A label task processing device, comprising: A processing instruction receiving module is configured to receive a label task processing instruction and obtain a label task processing list based on the label task processing instruction; An execution order generation module is configured to obtain a tag task to be executed from the tag task processing list, and generate a tag task execution order based on the tag task to be executed, wherein the generation of the tag task execution order based on the tag task to be executed includes: determining a tag task execution statement corresponding to the tag task to be executed; determining a parameter variable in the tag task execution statement based on a preset processing condition of the tag task to be executed; generating a tag task execution order for the tag task to be executed based on the parameter variable; wherein the preset processing condition is a scheduling frequency; and the parameter variable is an execution time; A task processing module is configured to determine a corresponding label processing model based on the label task execution form, and process the label task to be executed according to the label task execution form and the corresponding label processing model, wherein the label processing model is a processing logic for label data customized by a label developer, and the corresponding label processing model is determined based on the label task execution form, and the label task to be executed is processed according to the label task execution form and the corresponding label processing model, including: determining the label processing model called by the label task execution form, and processing the label task to be executed according to the label task execution form and the corresponding label processing model.

12. A method for generating visual labels based on artificial intelligence, comprising: receiving a visual label generation instruction, and generating a visual label generation task based on the visual label generation instruction; receiving a label task execution instruction for the visual label generation task, and acquiring a label task processing list based on the label task execution instruction; The visual label generation task is obtained from the label task processing list, and a label task execution list is generated based on the visual label generation task, wherein the label task execution list generated based on the visual label generation task includes: determining a label task execution statement corresponding to the visual label generation task; determining a parameter variable in the label task execution statement based on a preset processing condition of the visual label generation task; generating a label task execution list of the visual label generation task based on the parameter variable; wherein the preset processing condition is a scheduling frequency; and the parameter variable is an execution time; Determine the corresponding artificial intelligence model based on the label task execution form, and process the visual label generation task according to the label task execution form and the artificial intelligence model to generate the visual label, wherein the artificial intelligence model is a machine learning model that can perform secondary or multiple processing on basic video labels to generate visual labels, and determine the corresponding artificial intelligence model based on the label task execution form, and process the visual label generation task according to the label task execution form and the artificial intelligence model, including: determining the artificial intelligence model called by the label task execution form, and processing the visual label generation task according to the label task execution form and the artificial intelligence model.

13. A visual label generation device based on artificial intelligence, comprising: A generation instruction receiving module is configured to receive a visual label generation instruction and generate a visual label generation task based on the visual label generation instruction; A task list acquisition module is configured to receive a label task execution instruction for the visual label generation task, and acquire a label task processing list based on the label task execution instruction; An execution order generation module is configured to obtain the visual label generation task from the label task processing list, and generate a label task execution order based on the visual label generation task, wherein generating the label task execution order based on the visual label generation task includes: determining the label task execution statement corresponding to the visual label generation task; determining the parameter variables in the label task execution statement based on the preset processing conditions of the visual label generation task; generating the label task execution order of the visual label generation task based on the parameter variables; wherein the preset processing condition is the scheduling frequency; and the parameter variable is the execution time; A visual label generation module is configured to determine a corresponding artificial intelligence model based on the label task execution form, and process the visual label generation task according to the label task execution form and the artificial intelligence model to generate the visual label, wherein the artificial intelligence model is a machine learning model that can perform secondary or multiple processing on basic video labels to generate visual labels. The corresponding artificial intelligence model is determined based on the label task execution form, and the visual label generation task is processed according to the label task execution form and the artificial intelligence model, including: determining the artificial intelligence model called by the label task execution form, and processing the visual label generation task according to the label task execution form and the artificial intelligence model.

14. A computing device comprising: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions, wherein when the processor executes the computer-executable instructions, it implements the steps of the label task processing method described in any one of claims 1 to 10 or the artificial intelligence-based visual label generation method described in claim 12.

15. A computer-readable storage medium storing computer instructions, which, when executed by a processor, implement the steps of the label task processing method described in any one of claims 1 to 10 or the artificial intelligence-based visual label generation method described in claim 12.

Images