A Multiclass Boosting Framework for Achieving Fast and Provable Adversarial Robustness

TL;DR

This paper introduces a multiclass boosting framework that provides fast, provable adversarial robustness for neural networks, outperforming existing methods both in effectiveness and training efficiency.

Contribution

The authors develop a novel boosting-based approach with theoretical guarantees for adversarial robustness, demonstrating superior empirical performance and reduced training time.

Findings

Outperforms state-of-the-art adversarial robustness methods

Achieves robustness with less training time

Provides theoretical guarantees under certain conditions

Abstract

Alongside the well-publicized accomplishments of deep neural networks there has emerged an apparent bug in their success on tasks such as object recognition: with deep models trained using vanilla methods, input images can be slightly corrupted in order to modify output predictions, even when these corruptions are practically invisible. This apparent lack of robustness has led researchers to propose methods that can help to prevent an adversary from having such capabilities. The state-of-the-art approaches have incorporated the robustness requirement into the loss function, and the training process involves taking stochastic gradient descent steps not using original inputs but on adversarially-corrupted ones. In this paper we propose a multiclass boosting framework to ensure adversarial robustness. Boosting algorithms are generally well-suited for adversarial scenarios, as they were classically designed to satisfy a minimax guarantee. We provide a theoretical foundation for this methodology and describe conditions under which robustness can be achieved given a weak training oracle. We show empirically that adversarially-robust multiclass boosting not only outperforms the state-of-the-art methods, it does so at a fraction of the training time.