Tag Requirements for Different Data Types

A correct set of tags should be specified when adding files to the system folders so that the neural networks would process them correctly. You can always find the full information on tags and how to work with them in our documentation for developers.

Tags for Video Files

The following tag types should be specified in the system for video files:

  • A tag to identify the data type of the video:
    • video_selfie.
  • A tag to identify the orientation of the video:
    • orientation_portrait — portrait orientation;
    • orientation_landscape — landscape orientation.
  • A tag to identify the action on the video (this list is only an example, please consider consulting with technical documentation):
    • video_selfie_left — head turn to the left;
    • video_selfie_right — head turn to the right;
    • video_selfie_down — head tilt downwards;
    • video_selfie_high — head raise up;
    • video_selfie_smile — smile;
    • video_selfie_eyes — blink;
    • video_selfie_scan — scanning.

Example of a correct tag set for the turning head to the left gesture:

JSON

                "video1": [
      "video_selfie",
      "video_selfie_left",
      "orientation_portrait"
    ]


            

Tags for Photo Files

The following tag types should be specified in the system for photo files:

  • A tag for selfies:
    • photo_selfieto identify the image type as selfie.
  • Tags for photos/scans of ID cards:
    • photo_idto identify the image type as ID;
    • photo_id_frontfor the photo of the ID front side;
    • photo_id_backfor the photo of the ID back side.

Example of a correct tag set for the Selfie photo file:

JSON

                
"photo1": [
        "photo_selfie"
    ]


            

Example of the correct tag set for a photo file with the face side of an ID card:

JSON

                "photo1": [
        "photo_id",
        "photo_id_front"
    ]


            

Example of the correct set of tags for a photo of the back of an ID card:

JSON

                "photo1": [
        "photo_id",
        "photo_id_back"
    ]