fahadshamshad / Clip2Protect

[CVPR 2023] Official repository of paper titled "CLIP2Protect: Protecting Facial Privacy using Text-Guided Makeup via Adversarial Latent Search".
https://fahadshamshad.github.io/Clip2Protect/
98 stars 12 forks source link

How can we use your method when the user do not want to change anything attribute on his face #1

Closed lizi123321123 closed 1 year ago

lizi123321123 commented 1 year ago

hello! Your paper is very interesting. But I have some question about application scenario. When user do not want to change anything attribute about his face, how your method can work to protect privacy. I very hope that I can get chance to communicate with you.

fahadshamshad commented 1 year ago

Hello, if you do not wish to alter facial attributes, you can set the weightage of the clip loss to 0. This ensures no changes are made while still offering protection. However, it might generate minor visual distortions or artifacts.