# Robustifying $\\ell\_\\infty$ Adversarial Training to the Union of Perturbation Models

Source: [Robustifying $\\ell\_\\infty$ Adversarial Training to the Union of Perturbation Models](<https://www.sota2.com/research/paper/68fa6a291e209131cea26773>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-28T13:56:48.324Z

Authors: Ameya D. Patil, Michael Tuttle, Alexander G. Schwing, Naresh R. Shanbhag

Published: 2021-05-31T05:18:42.000Z

## Abstract

Classical adversarial training (AT) frameworks are designed to achieve high adversarial accuracy against a single attack type, typically $\\ell\_\\infty$ norm-bounded perturbations. Recent extensions in AT have focused on defending against the union of multiple perturbations but this benefit is obtained at the expense of a significant (up to $10\\times$) increase in training complexity over single-attack $\\ell\_\\infty$ AT. In this work, we expand the capabilities of widely popular single-attack $\\ell\_\\infty$ AT frameworks to provide robustness to the union of ($\\ell\_\\infty, \\ell\_2, \\ell\_1$) perturbations while preserving their training efficiency. Our technique, referred to as Shaped Noise Augmented Processing (SNAP), exploits a well-established byproduct of single-attack AT frameworks -- the reduction in the curvature of the decision boundary of networks. SNAP prepends a given deep net with a shaped noise augmentation layer whose distribution is learned along with network parameters using any standard single-attack AT. As a result, SNAP enhances adversarial accuracy of ResNet-18 on CIFAR-10 against the union of ($\\ell\_\\infty, \\ell\_2, \\ell\_1$) perturbations by 14%-to-20% for four state-of-the-art (SOTA) single-attack $\\ell\_\\infty$ AT frameworks, and, for the first time, establishes a benchmark for ResNet-50 and ResNet-101 on ImageNet.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2105.14710>)

[PDF](<https://arxiv.org/pdf/2105.14710.pdf>)
