This is for releasing the source code of the paper BruSLeAttack: Query-Efficient Score-Based Sparse Adversarial Attack
Poster: Poster
Artifacts: Artifacts.
These clean and perturbed samples & artifacts can be tested on Google Cloud Vision (Test them for free!!!)
Cite our research:
@inproceedings{Vo2024,
title = {BruSLeAttack: A Query-Efficient Score-Based Black-Box Sparse Adversarial Attack},
year = {2024},
booktitle = {International Conference on Learning Representations (ICLR)},
author = Viet Quoc Vo and Ehsan Abbasnejad and Damith C. Ranasinghe},
}
- Add the testing code.