Adversarial robustness vs. model compression, or both?
Shaokai Ye, Kaidi Xu, et al.
ICCV 2019
Motivated by the recent discovery that the interpretation maps of CNNs could easily be manipulated by adversarial attacks against network interpretability, we study the problem of interpretation robustness from a new perspective of Rényi differential privacy (RDP). The advantages of our Rényi-Robust-Smooth (RDP-based interpretation method) are three-folds. First, it can offer provable and certifiable top-k robustness. That is, the top-k important attributions of the interpretation map are provably robust under any input perturbation with bounded ℓd-norm (for any d≥1, including d=∞). Second, our proposed method offers ∼12% better experimental robustness than existing approaches in terms of the top-k attributions. Remarkably, the accuracy of Rényi-Robust-Smooth also outperforms existing approaches. Third, our method can provide a smooth tradeoff between robustness and computational efficiency. Experimentally, its top-k attributions are twice more robust than existing approaches when the computational resources are highly constrained.
Shaokai Ye, Kaidi Xu, et al.
ICCV 2019
Chia-yi Hsu, Jia You Chen, et al.
ICASSP 2025
Akshay Mehra, Bhavya Kailkhura, et al.
NeurIPS 2020
Sijia Liu, Bhavya Kailkhura, et al.
NeurIPS 2018