Personal Data Anonymization

Module Description

Personal Data Anonymization can blur faces and license plates in images and videos. It is a privacy protection tool that helps you to comply with data protection regulations.

Module ID: personal_data_anonymization

Module Parameters

Name	Type	Default	Description
pipes	list	["face", "license_plate"]	The detection pipeline to apply. Possible values are: face, license_plate.
threshold_face_detection	number	0.6	The minimum confidence threshold for face detection (range from 0.0 to 1.0).
threshold_face_recognition	number	1.0	Thresholding the distance, how similar (in terms of euclidean distance) the recognized identity is to the closest sample in the training data (range from 0.0 to 5.0).
threshold_license_plate	number	0.7	The minimum confidence threshold for license plate detection (range from 0.0 to 1.0).
blur_mode_face	str	whitelist	The blur mode when face recognition is applied (don't blur known faces, blur known faces, blur all). Possible values are: whitelist, blacklist, all.
padding_ratio	number	0.2	The padding ratio for the blur area (bounding-box) of the detected object (range from 0.0 to 1.0).
dump_overlays	boolean	false	Dump the overlays as separate image files returned as job artifacts.

Pre-trained models

Example

Send the following JSON as request body via POST to the /jobs/ endpoint:

{
  "sources": [
    "{url-to-your-image}"
  ],
  "modules": {
    "personal_data_anoymization": {
      "pipes": ["face"]
    }
  }
}

When requesting the job via GET on the /jobs/{JOB_ID}/ endpoint, the response looks like this:

coming soon