A Fast Saddle-Point Dynamical System Approach to Robust Deep Learning

TL;DR

This paper introduces a novel saddle-point dynamical system algorithm for robust deep learning, significantly reducing computational costs while achieving high robustness against adversarial attacks.

Contribution

It proposes a new discrete-time dynamical system-based algorithm for robust training that converges under convexity assumptions and demonstrates improved efficiency and robustness.

Findings

Algorithm converges asymptotically to the robust optimal solution.

Achieves significant robustness improvements over state-of-the-art methods.

Reduces computational cost of robust training on benchmark datasets.

Abstract

Recent focus on robustness to adversarial attacks for deep neural networks produced a large variety of algorithms for training robust models. Most of the effective algorithms involve solving the min-max optimization problem for training robust models (min step) under worst-case attacks (max step). However, they often suffer from high computational cost from running several inner maximization iterations (to find an optimal attack) inside every outer minimization iteration. Therefore, it becomes difficult to readily apply such algorithms for moderate to large size real world data sets. To alleviate this, we explore the effectiveness of iterative descent-ascent algorithms where the maximization and minimization steps are executed in an alternate fashion to simultaneously obtain the worst-case attack and the corresponding robust model. Specifically, we propose a novel discrete-time dynamical system-based algorithm that aims to find the saddle point of a min-max optimization problem in the presence of uncertainties. Under the assumptions that the cost function is convex and uncertainties enter concavely in the robust learning problem, we analytically show that our algorithm converges asymptotically to the robust optimal solution under a general adversarial budget constraints as induced by $\ell_p$ norm, for $1\leq p\leq \infty$. Based on our proposed analysis, we devise a fast robust training algorithm for deep neural networks. Although such training involves highly non-convex robust optimization problems, empirical results show that the algorithm can achieve significant robustness compared to other state-of-the-art robust models on benchmark data sets.