OPTML-Group / Diffusion-MU-Attack

The official implementation of ECCV'24 paper "To Generate or Not? Safety-Driven Unlearned Diffusion Models Are Still Easy To Generate Unsafe Images ... For Now". This work introduces one fast and effective attack method to evaluate the harmful-content generation ability of safety-driven unlearned diffusion models.
MIT License
59 stars 3 forks source link

Could you provide unlearned models for objects? #6

Closed TianyunYoung closed 9 months ago

TianyunYoung commented 9 months ago

Hi~ The work is awesome! I am trying to reproduce the results. I notice that the provided models in the google drive link does not include unlearned models for objects such as church, parachute, tench, and garbage truck. Could you please provide these models? Thank you very much.

damon-demon commented 9 months ago

Hello!

Thank you very much for expressing interest in our project! We appreciate your enthusiasm. For your convenience, the object-unlearned checkpoints are accessible via the following Google Drive link: https://drive.google.com/file/d/11FNKE_zzm8NMvUQn2_UPWzZRQFd4Frgb/view?usp=sharing.

Wishing you a wonderful day! ☀️

TianyunYoung commented 9 months ago

Thank you very much