Adversarial Robustness Curves
This work addresses the challenge of trust in automated systems due to adversarial examples, but it is incremental as it builds on existing robustness research.
The paper tackles the problem of analyzing adversarial robustness independently of specific thresholds and norms by proposing robustness curves as a more general framework, and investigates how these curves depend on the chosen norm.
The existence of adversarial examples has led to considerable uncertainty regarding the trust one can justifiably put in predictions produced by automated systems. This uncertainty has, in turn, lead to considerable research effort in understanding adversarial robustness. In this work, we take first steps towards separating robustness analysis from the choice of robustness threshold and norm. We propose robustness curves as a more general view of the robustness behavior of a model and investigate under which circumstances they can qualitatively depend on the chosen norm.