Tasks section#

The “Tasks” is intended for creating, deleting, and displaying tasks, downloading search results by events and persons. Export to a file is implemented in the Interface in the form of a task;

General view of the “Tasks” section is shown below (Figure 51).

The “LUNA PLATFORM. Deferred tasks” tab contains the following elements:

task counter (1);
“Cross-matching” button — button for creating a task for cross-matching lists of faces;
“Export faces” button — button for creating a task to export faces and information on them;
“Export events” button — button for creating a task to export events and information on them;
“Batch processing” button — button for creating a task for batch processing of photo image archives according to a specific policy;
“Batch import” button — button for creating a task for batch import of photo image archive into the list;
“Batch identification” — button for creating a task for identifying an archive of photo images of references with candidates (faces or events with faces);
“Deleting faces from the list” — button for creating a task for removing persons from the selected list;
table of tasks:
“ID” — task identifier;
“Description” — user who created the task;
“Type” — task type (cross-matching, export, batch processing, batch import, batch identification);
“Date created” — date and time of task creation;
“Expiration date” — date and time of completion of the task;
“Status” — task progress state;
— button to stop the task (appears if the task status is “In progress”);
— button for downloading the result of the task (2);
— button for deleting a task (3);
the number of tasks displayed on the page is set by the switch in the lower right corner of the page. There can be 10, 25, 50 or 100 tasks in total on one page (4).

The status changes during the task execution. In total, 4 statuses are applied to tasks in the Interface:

— task is being performed;
“Collecting results” — collecting the results of the task;
— task completed;
— an error occurred while executing a task.

The process of creating tasks and the values of the specified parameters are described below. If you need to go back to the task list page during creating a task, press the Esc key on your keyboard.

Configure notifications about task status using the "callbacks" functionality. Notifications will be sent to the external system at the specified URL. The notification settings block opens after filling in the required fields to create a task (Table 25).

Table 25. Notification settings in the task creation form

Parameter	Description	Default value
Add callback
Type	Protocol type when creating a notification. Users can receive notifications in the interface via http and Telegram	HTTP
URL	Address of the external system where the notification will be sent	-
Authorization type	Selecting the type of authorization into an external system and setting up authorization data. The basic type of authorization requires specifying login and password to enter an external system	Basic
Timeout (seconds)	Maximum time to wait for a request to complete	60
Request body format	Data interchange format: JSON or MessagePack	application/json
HTTP Headers	HTTP Request Headers	-

Creating a cross-matching Task#

To create a task for cross-matching lists of faces, click on the “Cross-matching” button (Figure 51). A general view of the form for creating a cross-matching task is shown below (Figure 52).

The “Cross-matching” form contains the following elements:

“List”* — selection of a list for comparison;
“Find matches in”* — selection of a list for comparison;
“The maximum number of similar ones” — maximum number of similar candidates (the default is 3);
“Minimum similarity threshold, %” — the lowest score of similarity in percentage between candidates that the Interface accepts as a possible match (the default is 50).

* Required field

Fill in all the required parameters and click on the “Create task” button (or Enter key on your keyboard)..

Resource-intensive tasks can take a while. In the pop-up window (Figure 53), you must confirm the action — click on the “Ok” button or cancel the action by clicking on the “Cancel” button (Esc key on keyboard)..

Figure 53. Confirmation of the creation of a cross-matching task

After successfully creating a cross-matching task, the message “Task for cross-matching has been created” will appear in the upper right corner of the screen (Figure 54).

Figure 54. Confirmation of successful creation of a cross-matching task

Creating an export faces task#

To create a task to export faces and information on them, click on the “Export faces” button (Figure 51). A general view of the form for creating an export task is shown below (Figure 55).

Description of the parameters of the “Export faces” form is given in Table 26.

Table 26. Export faces task parameters

Parameter	Description	Default value
Data upload parameters
List	Specifies the list for export	-
User data	Indicates face data (up to 128 characters)	-
Face IDs	Specifies the values of identifiers of faces in LUNA PLATFORM 5 in UUID format	-
Face external IDs	Specifies the values of third-party external identifiers	-
Date from	Specifies the lower limit of the period of creation of faces or events in LUNA PLATFORM 5	-
Date to	Specifies the upper limit of the period of creation of faces or events in LUNA PLATFORM 5	-
ID of the first face	Specify the values of the identifier of the first face from the exported faces	-
ID of the last face	Specify the values of the identifier of the last face from the exported faces	-
Additional parameters
Columns in the report—selecting table columns to be included in the file upon export, and indication the order in which they are located	Face ID User data External ID Time Avatar Event ID Lists	On On On On On On Off
Save face images	Enabling this option allows you to upload face images into the archive with the .csv report	Off
Type of biometric template	Specifies the biometric template of which objects will be exported—faces or bodies	Faces
Delimiter for .csv	A special character that will be used in the file with export results to divide text into columns	,

Fill in all the required parameters and click on the “Create task” button.

Resource-intensive tasks can take a while. In the pop-up window (Figure 56), you must confirm the action — click on the “Ok” button or cancel the action by clicking on the “Cancel” button (Esc key on keyboard)..

Figure 56. Confirmation of the creation of an export task

After successfully creating an export task, the message “Export task has been successfully created” will appear in the upper right corner of the screen (Figure 57).

Figure 57. Confirmation of successful creation of an export task

Creating an export events task#

To create a task to export objects events and information on them, click on the “Export events” button (Figure 51). A general view of the form for creating an export events task is shown below (Figure 58).

Description of the parameters of the “Export events” form is given in Table 27.

Table 27. Export events task parameters

Parameter	Description	Default value
Data upload parameters
General data about the event
Source	Specifies the name of the event source	-
Event IDs	Specifies the values of the event identifiers in UUID format for performing an accurate search separated by commas	-
User data	Indicates face data (up to 128 characters)	-
Event external IDs	Specifies the values of third-party external identifiers separated by commas	-
Face IDs	Specifies the values of identifiers of faces in UUID format separated by commas	-
Similarity	A value from 0 to 1 is specified	-
Tags	Specifies a tag or tags separated by commas	-
Handling policies	Specifies policy name, it is possible to specify several values	-
Date from	Specifies the lower limit of the period of creation of events	-
Date to	Specifies the upper limit of the period of creation of events	-
Advanced event filters
End date from	Indicates the lower limit of the event end period	-
End date to	Indicates the upper limit of the event end period	-
ID of the first event	Indicate the value of the identifier of the first event from the exported events	-
ID of the last event	Indicates the value of the identifier of the last event from the exported events	-
Track IDs	Specifies the values of the track identifiers in the UUID format separated by commas	-
Face attributes
Gender	Specifies male/female gender	-
Age category	Specifies the age range	-
Emotions	Specifies emotions	-
Medical mask	Detection of the presence/absence of a medical mask, mouth occlusion	-
Liveness	Specifies the result of checking for the presence of a living person in the frame
Body attributes
Gender by body	Specifies the female, male, undefined gender	-
Age category by body	Specifies the age range by body image	-
Headwear	Specifies headdress	-
Upper body colors	Specifies top clothing color	-
Sleeve	Specifies sleeve length	-
Headwear color	Specifies headdress color	-
Lower body colors	Specifies bottom clothing color	-
Lower body type	Specifies bottom clothing type	-
Shoes color	Specifies shoe color (only for “Identify among events);
Backpack	Specifies backpack presence	-
Best match data
Face IDs	Specifies the values of identifiers of faces 5 in UUID format	-
Face external IDs	Specifies the values of third-party external identifiers	-
Label	Name of the label parameter (the rule by which comparison occurs);	-
Geoposition	“District”; “Area”; “City”; “Street”; “House number”;
Advanced geoposition filters	“Longitude (-180…180)”; “Accuracy (0…90)”; “Latitude (-90…90)”; “Accuracy (0…90)”
Other
Add filter by meta**	Allows you to fill in a set of blocks to create a filter by the "meta" field. The number of meta filters is unlimited. The following blocks are required to be filled in when creating a filter by meta: Key—the full path to the required meta field connected with events. Value—any valid value for the meta field. Type—selection of the data type stored in this meta field. The data type displays available operators and converts values into the desired data type when sent to the API. Operator—select operators for a given data type from the list. Operator type depends on the selected data type
Additional parameters
Columns in the report—selecting table columns to be included in the file upon export, and indication the order in which parameters are located in the report	Event data DFace data from the event Body data from the event Match data Geo position	Off Off Off Off On
Save face images	Enabling this option allows you to upload face images into the archive with the .csv report	Off
Type of biometric template	Specifies the biometric template of which objects will be exported—faces or bodies	Faces
Delimiter for .csv	A special character that will be used in the file with export results to divide text into columns	,

Fill in all the required parameters and click on the “Create task” button or the Enter key on your keyboard.

Resource-intensive tasks can take a while. In the pop-up window, confirm the action — click on the “Ok” button or cancel the action by clicking on the “Cancel” button (Esc key on keyboard).

After successfully creating an export task, the message “Export task has been successfully created” will appear in the upper right corner of the screen.

Creating a batch processing task#

The batch processing task allows user to process several photos using a specified policy.

To create a task for batch processing of photo image archives according to a specific policy, click on the “Batch processing” button (Figure 51). The general view of the form for creating a batch processing task is shown below (Figure 59).

Figure 59. Form for creating a batch processing task

By default the “Batch Processing” form contains the following elements:

“Data source type” — selection of the source type of the loaded data;
“Description” — description of the task;
“Handling policy”* — selection of a policy (required);

The resource can accept five types of sources with images for processing:

File;
ZIP;
S3;
Network disk;
FTP;
Samba.

Additional options appear depending on the selected data source type.

To quickly download a ZIP archive from your local machine without additional options, select “File” as the data source type. Then upload or drag-and-drop the archive with photo images in the field for uploading data.

Download file requirements:

*.zip file format;
there can be one or more people on the image (depends on policy settings);
the image must contain a person's face or body;
images must be located immediately inside the archive, and not in a folder inside the archive;
the archive size is set using the ARCHIVE_MAX_SIZE parameter in the config.py configuration file of the Tasks component, the default size is 100 GB (for details, see “VisionLabs LUNA PLATFORM 5. Administrator manual”).

When choosing a ZIP archive as image source for the batch processing task, the following parameters can be set:

“File URL”* — URL address of the archive with images, the default archive size is 100 GB;;
“Archive password” — a password for the transferred archive protection;
“File key prefix” — a file key prefix that can be used to load images from a specific directory, for example, "2022/January";;
“File key postfix” — file key postfix that can be used to upload images with a specific extension;
“Whether to estimate images from ZIP archive subdirectories recursively?” switch — allows you to recursively receive images from subdirectories.
“Input image type” — selection of the type of image that is input in the batch processing task — "Raw image", "Face warped image", "Body warped image".

When choosing an S3-like storage as an image source for the batch processing task, the following parameters can be set:

“Storage endpoint” — only when specifying the bucket name;
“Bucket name”* — Access Point ARN / Outpost ARN;
“File key prefix” — file key prefix. It can be used to load images from a specific folder, such as "2022/January";
“Bucket region” — only when specifying the bucket name;
“Public access key”* — public key for setting up authorization;
“Secret access key”* — secret key for setting up authorization;
“Signature version” — signature "s3v2" / "s3v4";
“Whether to estimate images from bucket subdirectories recursively?” — possibility to recursively download images from nested bucket folders;
“Whether to save image origin?” — saving original images in the LUNA PLATFORM 5 database.

It is also possible to select the type of transferred images. For more information about working with S3-like repositories, see AWS User Guide.

When choosing a network disk as an image source for the batch processing task, the following parameters can be set:

“Path to directory with images”* — absolute path to the directory with images in the container (required);
“File key prefix” — a file key prefix that can be used to load images from a specific directory;
“File key postfix” — file key postfix that can be used to upload images with a specific extension;
“Whether follow file system links?” — enable/disable of symbolic links processing.

As in the batch processing task using S3-like storage as image source, it is possible to recursively receive images from nested directories, and to select the type of transferred images.

When choosing a FTP server as an image source for the batch processing task, the following parameters can be set:

“Server host”* — FTP server IP address or hostname;
“Port” — FTP server port;
“FTP sessions” — maximum number of allowed sessions on the FTP server;
“Server user” and “Server password” — authorization parameters.

* Required field

As in the batch processing task using network disk as image source, it is possible to set the path to the directory with images, recursively receive images from nested directories, select the type of transferred images, and specify the prefix and postfix.

When choosing a Samba as an image source for the batch processing task, the parameters are similar to those of an FTP server, except for the "max_sessions" parameter. Also, if authorization data is not specified, the connection to Samba will be performed as a guest.

Fill in all the required parameters and click on the “Create task” button or the Enter key on your keyboard. Resource-intensive tasks can take a while. In the pop-up window (Figure 60), you must confirm the action — click on the “Ok” button or cancel the action by clicking on the “Cancel” button.

Figure 60. Confirmation of the creation of a batch processing task

After successfully creating a batch processing task, the message “Batch processing task has been successfully created” will appear in the upper right corner of the screen (Figure 61).

Figure 61. Confirmation of successful creation of a batch processing task

Creating a batch import task#

The batch import task allows you to batch import faces from photos into a specified list.

To create a task for batch import of photo image archive into the list, click on the “Batch import” button (Figure 51). The general view of the form for creating a batch import task is shown below (Figure 62).

Figure 62. Form for creating a batch import task

The “Batch import” form contains the following elements:

field for uploading an archive with photographs* — it is possible to upload archives in *.ZIP format (required);
“List” — selection of a list (required);
“Add a photo to the list only if it complies with the ISO/IEC standard” — the photo will be added to the list only after passing the ISO/IEC 19794-5:2011 verification.
— button for deleting the loaded archive — button for deleting the loaded archive.

* Required to be filled out.

Download file requirements:

*.ZIP file format;
there can be one or more people on the image (depends on policy settings);
the image must contain a person's face;
images must be located immediately inside the archive, and not in a folder inside the archive;
the archive size is set using the ARCHIVE_MAX_SIZE parameter in the config.py configuration file of the Tasks component, the default size is 100 GB (for details, see “VisionLabs LUNA PLATFORM 5. Administrator manual”).

Fill in all the required parameters and click on the “Create task” button or the Enter key on your keyboard.

Resource-intensive tasks can take a while. In the pop-up window (Figure 63), you must confirm the action — click on the “Ok” button or cancel the action by clicking on the “Cancel” button.

Figure 63. Confirmation of the creation of a batch import task

After successfully creating a batch import task, the message “Batch import task has been successfully created” will appear in the upper right corner of the screen (Figure 64).

Figure 64. Confirmation of successful creation of a batch import task

Creating a batch identification task#

To create a task for identifying an archive of photo images (faces or events with faces), click on the “Batch identification” button (Figure 51). The general view of the form for creating a batch identification task is shown below (Figure 65).

The “Batch Identification” form contains the following elements:

field for uploading an archive with photographs in *.ZIP format (required);
—button for deleting the loaded archive;
“Identify among”—look for matches among “Faces” or “Events”;
“Filters” block—settings for user identification. A description of the parameters of the “Filters” block, depending on the selected object for identification, is presented below (Table 28 and Table 29);
“Additional filter parameters” block—general parameters for identification among faces and events:
“Similarity threshold, %”—the lowest percentage similarity score between candidates that the Interface accepts as a possible match (default: 80).
“Number of records (from 1 to 100)”—the number of lines with matches with a limit of 100 lines (default: 3).

Table 28. “Filters” block parameters of the batch identification task when searching for a match among faces

Name	Description
List	List name
Comma-separated Face IDs	Face ID from the list
User data	Information linked to the person from the database
Comma-separated external face IDs	External ID of persons face
Date from	Specifies the lower limit of the period of creation of faces or events in LUNA PLATFORM 5
Date to	Specifies the upper limit of the period of creation of faces or events in LUNA PLATFORM 5

Table 29. “Filters” block parameters of the batch identification task when searching for a match among events

Name	Description
Source	List of available event sources
User data	Information linked to the person from the database
Age category	Age group: below 18; from 18 to 44; from 45 to 60; above 60
Gender	Female; Male Not estimated
Emotion	Anger; Sadness; Neutral; Disgust; Fear; Happiness; Surprise. Its possibly to select multiple emotions.
Medical mask	Detection of the presence/absence of a medical mask or mouth occlusion. Missing; Medical mask; Occluded. Its possibly to select multiple variants.
Creation date from	Specifies the lower limit of the period of creation of faces or events in LUNA PLATFORM 5
Creation date to	Specifies the upper limit of the period of creation of faces or events in LUNA PLATFORM 5
Comma-separated event IDs	Event ID of detection and attribute retrieval
Comma-separated external event IDs	External event ID
Comma-separated Face IDs	Face ID from events that are created in LUNA PLATFORM 5 as a result of a detection event and attribute extraction
Similarity	The lower threshold on the similarity if the person was identified
Handling policies	Handling Policy ID
Comma-separated track IDs	Specifies the values of the track identifiers in LUNA PLATFORM 5 in the UUID format
Comma-separated tags	Specifies a tag or tags
Gender by body	female male undefined
Headwear	Specifies headdress
Sleeve	Specifies sleeve length
Lower body colors	Specifies bottom clothing color
Shoes color	Specifies shoe color (only for “Identify among events);
Age category by body	Specifies the age range below 18; from 18 to 44; from 45 to 60; above 60
Upper body colors	Specifies top clothing color
Headwear color	Specifies headdress color
Lower body type	Specifies bottom clothing type
Backpack	Specifies backpack presence
Location	“District”; “Area”; “City”; “Street”; “House number”; “Longitude (-180…180)”; “Accuracy (0…90)”; “Latitude (-90…90)”; “Accuracy (0…90)”

To upload an archive with photo images of faces to be identified, click on in the “References” section and specify the path to the archive on the local computer.

Download file requirements:

*.ZIP file format;
there can be one or more people on the image (depends on policy settings);
the image must contain a person's face;
images must be located immediately inside the archive, and not in a folder inside the archive;
the archive size is set using the ARCHIVE_MAX_SIZE parameter in the config.py configuration file of the Tasks component, the default size is 100 GB (for details, see “VisionLabs LUNA PLATFORM 5. Administrator manual”).

Fill in all the necessary parameters and click the “Create task” button or the Enter key on your keyboard..

Resource-intensive tasks can take a while. In the pop-up window (Figure 66), you must confirm the action — click on the “Ok” button or cancel the action by clicking on the “Cancel” button.

Figure 66. Confirmation of the creation of a batch identification task

After successfully creating a batch identification task, the message “Batch identification task has been successfully created” will appear in the upper right corner of the screen (Figure 67).

Figure 67. Confirmation of successful creation of a batch identification task

Creating a task for deleting faces from the list#

The task of removing persons from the list (Cleanup task) allows you to select faces based on specific parameters and then remove them from the selected list.

To create a Cleanup task, click on the “Deleting faces from the list” button (Figure 51). The general view of the window for creating a task for batch import is shown below (Figure 68).

Figure 68. Window for creating a “Deleting faces from the list” task

The “Deleting faces from the list” window contains the following elements:

“Description”—a field for adding an explanatory note to the task;
“Store results” checkbox—if enabled, the results of the task will be saved in the Image Store service storage.
“Delete samples” checkbox—if enabled, wrapped images obtained after detecting faces from the list will be deleted;
“List”*—select a list from which faces will be removed;
“Information”—a field for specifying information about persons from the list. Allows you to remove only a few people from the list, for example, those for whom the same information is provided.
“Delete data after”—the lower included threshold value of the face creation time;
“Delete data before”—the upper excluded threshold value of the face creation time.

* Required to be filled out.

Fill in all the required parameters and click on the “Create task” button or the Enter key on your keyboard.

Resource-intensive tasks may take some time to complete. Confirm the action in the pop-up window—click the “Ok” button or cancel the action using the “Cancel” button (Figure 69).

Figure 69. Confirmation of creating a “Deleting faces from the list” task

After successfully creating a task for removing persons from the list, the message “Cleanup task has been successfully created” appears on the screen (Figure 70).

Figure 70. Confirmation of successful creation of the “Deleting faces from the list” task

Viewing the results of a task#

Viewing the results of a task is performed by pressing the button in the line (3 in the Figure 51). For viewing the *.ZIP archive for export tasks, the *.csv file for cross-matching tasks, the *.json file for batch processing, batch import and batch identification tasks (where * is the task ID) will be loaded.

The downloaded *.csv file contains a table with the export parameters selected in "Creating an export task" section (Figure 71) or with the results of cross-matching (Figure 72).

Task deleting#

Deleting a task is performed by clicking the button in the line (4 in the Figure 51).

In the pop-up window (Figure 73), you must confirm the action — click on the “Delete” button or cancel the action by clicking on the “Cancel” button. After the successful deletion, a corresponding notification appears.

Figure 73. Confirmation of deletion of the task