Skip to content

[ACM MM 2025] ICAS: Detecting Training Data from Autoregressive Image Generative Models

License

Notifications You must be signed in to change notification settings

Chrisqcwx/ImageAR-MIA

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

ICAS: Detecting Training Data from Autoregressive Image Generative Models

ACM MM 2025

arxiv

Usage

Get the VAR repository:

git clone git@github.com:FoundationVision/VAR.git
rm ./VAR/README.md
mv ./VAR/* .
rmdir ./VAR

Modify the checkpoint and imagenet folder in line 6-7 and run membership inference:

# run baseline
python baseline.py
# run ICAS
python icas.py

Citation

@inproceedings{yu2025icas,
author = {Yu, Hongyao and Qiu, Yixiang and Yang, Yiheng and Fang, Hao and Zhuang, Tianqu and Hong, Jiaxin and Chen, Bin and Wu, Hao and Xia, Shu-Tao},
title = {ICAS: Detecting Training Data from Autoregressive Image Generative Models},
year = {2025},
booktitle = {Proceedings of the 33rd ACM International Conference on Multimedia},
pages = {11209–11217},
numpages = {9},
series = {MM '25}
}

About

[ACM MM 2025] ICAS: Detecting Training Data from Autoregressive Image Generative Models

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages