Moderation Overview

AI Tools

Moderation Requests

When setting up a moderation request you will need to provide the following:

Request

Description

URL of the content

The location of the content to be moderated. This needs to be publicly accessible / accessible to the Trust & Safety tool

Content type

The type of content to be checked. Currently supported types are: VIDEO, LIVE_VIDEO, IMAGE

Profile IDs that should be in the content

The profiles of the users that you are expecting to see in the video. These profiles will be checked against the faces found in the content

Age estimation threshold

The threshold for the age estimation check. Any unknown faces that are age estimated as below the threshold will trigger an UNDERAGE notification to be sent

Unknown Faces Flag

The flag to set if you want to be notified about any unknown faces found in the content. If set to TRUE, any faces found in the video that do not match the profiles to be checked will trigger a UNKNOWN_FACE notification to be sent

Banned Profiles Flag

The flag to set if you want to be notified about any banned faces found in the content. If set to TRUE, any unknown faces found in the content will be compared against the banned profiles of your account to check for matches. Any matches found will trigger a BANNED_PROFILE notification to be sent

Notifications configuration

Configure how you want to receive notifications for the moderation request. Currently we support notifications via Websocket, Slack or via a Webhook


Moderation Notifications

As the content gets moderated, notifications will be sent to you. These provides you with information for further review. The notifications include:

Notification

Description

Type

The notification type will be one of the following:

  • KNOWN_FACE

  • BANNED_FACE

  • UNKNOWN_FACE

  • AGE_ESTIMATION (underage)

Moderation ID

The ID for the moderation request

Content Type

The content type will be one of the following:

  • VIDEO

  • LIVE_VIDEO

  • IMAGE

Content URL

The URL of the content being moderated

Content ID

The ID of the content being moderated

Check ID

The ID of the check

Timestamp

The timestamp of the frame where the face was found

Coordinates

Coordinates of the face found that triggered the notification