ZeroMark

This is the official implementation of our paper ZeroMark: Towards Dataset Ownership Verification without Disclosing Watermark, accepted by NeurIPS 2024. This research project is developed based on Python 3 and Pytorch, created by Junfeng Guo.

Dependencies

Our code is implemented using Torch. Following packages are required.

PyTorch => 1.6.*
torchvision > 0.5.*
Our code is tested on Python 3.8.3

To generate the boundary gradients for samples belong to (target) class 0 uisng samples from (ori) class 1, run:

python zeromark.py --ori 1 --target 0

to get the normalized boundary gradient similarity by:

python ga.py

Name		Name	Last commit message	Last commit date
Latest commit History 4 Commits
models		models
saved_models		saved_models
utils		utils
.DS_Store		.DS_Store
.Rhistory		.Rhistory
README.md		README.md
data.tar.gz		data.tar.gz
ga.py		ga.py
zeromark.py		zeromark.py