Get our free extension to see links to code for papers anywhere online!

Chrome logo  Add to Chrome

Firefox logo Add to Firefox


Improved Image Wasserstein Attacks and Defenses

Apr 26, 2020
J. Edward Hu, Adith Swaminathan, Hadi Salman, Greg Yang


Share this with someone who'll enjoy it:


Robustness against image perturbations bounded by a $\ell_p$ ball have been well-studied in recent literature. Perturbations in the real-world, however, rarely exhibit the pixel independence that $\ell_p$ threat models assume. A recently proposed Wasserstein distance-bounded threat model is a promising alternative that limits the perturbation to pixel mass movements. We point out and rectify flaws in previous definition of the Wasserstein threat model and explore stronger attacks and defenses under our better-defined framework. Lastly, we discuss the inability of current Wasserstein-robust models in defending against perturbations seen in the real world. Our code and trained models are available at https://github.com/edwardjhu/improved_wasserstein .

* Best paper award at ICLR Trustworthy ML Workshop 2020 


   Access Paper Source



Share this with someone who'll enjoy it: