Skip to content

Official implementation of the paper "NullFace: Training-Free Localized Face Anonymization"

License

Notifications You must be signed in to change notification settings

hanweikung/nullface

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

4 Commits
 
 
 
 
 
 

Repository files navigation

NullFace: Training-Free Localized Face Anonymization

arXiv

Our method obscures identity while preserving attributes such as gaze, expressions, and head pose (in contrast to Stable Diffusion Inpainting) and enables selective anonymization of specific facial regions.

Test set

For the quantitative comparisons against baseline methods in our paper, we selected:

  • 4,852 test subjects from the CelebA-HQ dataset
  • 4,722 test subjects from the FFHQ dataset

For each subject, we created corresponding segmentation masks to selectively control the visibility of the eye and mouth areas if desired.

Download

The list of selected test subjects and their corresponding segmentation masks are available for download at the Hugging Face Hub.

Code coming soon

About

Official implementation of the paper "NullFace: Training-Free Localized Face Anonymization"

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published