Minimal Cascade Gradient Smoothing for Fast Transferable Preemptive Adversarial Defense

TL;DR

This paper introduces MSPD, a fast, transferable preemptive defense against adversarial attacks that works without model access, significantly improves robustness and speed, and is supported by theoretical proofs.

Contribution

The paper proposes MSPD and MCGS, novel methods for preemptive adversarial defense that are fast, transferable, and effective without needing target model gradients.

Findings

MSPD improves robust accuracy by +5% across multiple models.

MSPD runs 28-1696 times faster than prior methods.

MSPD maintains robustness even under adaptive white-box attacks.

Abstract

Adversarial attacks persist as a major challenge in deep learning. While training- and test-time defenses are well-studied, they often reduce clean accuracy, incur high cost, or fail under adaptive threats. In contrast, preemptive defenses, which perturb media before release, offer a practical alternative but remain slow, model-coupled, and brittle. We propose the Minimal Sufficient Preemptive Defense (MSPD), a fast, transferable framework that defends against future attacks without access to the target model or gradients. MSPD is driven by Minimal Cascade Gradient Smoothing (MCGS), a two-epoch optimization paradigm executed on a surrogate backbone. This defines a minimal yet effective regime for robust generalization across unseen models and attacks. MSPD runs at 0.02s/image (CIFAR-10) and 0.26s/image (ImageNet), 28--1696 times faster than prior preemptive methods, while improving robust accuracy by +5% and clean accuracy by +3.7% across 11 models and 7 attacks. To evaluate adaptive robustness, we introduce Preemptive Reversion, the first white-box diagnostic attack that cancels preemptive perturbations under full gradient access. Even in this setting, MSPD retains a +2.2% robustness margin over the baseline. In practice, when gradients are unavailable, MSPD remains reliable and efficient. MSPD, MCGS, and Preemptive Reversion are each supported by formal theoretical proofs. The implementation is available at https://github.com/azrealwang/MSPD.