swheaton/fiftyone-media-anonymization-plugin

FiftyOne Plugin to anonymize your media by searching for any query string!

Media Anonymization Plugin

This plugin is a Python plugin that allows you to anonymize media in your FiftyOne dataset by generating a copy of the images where each detection bounding box is blurred out.

1. [Optional] Run zero-shot prediction model to generate detections field on the dataset. This can be done in the app easily with the Zero Shot Prediction Plugin
2. Choose an existing fiftyone.Detections field (or use field from #1) to apply the blur effect over each bounding box to.
3. Choose a path for the new images to be saved.
4. Choose to save to an alternate media field or clone to a new anonymized version of the dataset.
5. Execute operator, and enjoy your new anonymized FiftyOne app consumption!

Installation

fiftyone plugins download https://github.com/swheaton/fiftyone-media-anonymization-plugin

Refer to the main README for more information about managing downloaded plugins and developing plugins locally.

Usage

1. Launch the App:

   import fiftyone as fo
   import fiftyone.zoo as foz
   
   dataset = foz.load_zoo_dataset("quickstart")
   session = fo.launch_app(dataset)

2. Press ` or click the Browse operations action to open the Operators list

3. Select the blur_objects operator listed below!

Operators

blur_objects

Use this operator to selectively blur objects in images based on any fiftyone.Detections field in the dataset.