Efficient Preemptive Robustification with Image Sharpening

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Liang, Jiaming, Pun, Chi-Man
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915892490665984
author Liang, Jiaming
Pun, Chi-Man
author_facet Liang, Jiaming
Pun, Chi-Man
contents Despite their great success, deep neural networks rely on high-dimensional, non-robust representations, making them vulnerable to imperceptible perturbations, even in transfer scenarios. To address this, both training-time defenses (e.g., adversarial training and robust architecture design) and post-attack defenses (e.g., input purification and adversarial detection) have been extensively studied. Recently, a limited body of work has preliminarily explored a pre-attack defense paradigm, termed preemptive robustification, which introduces subtle modifications to benign samples prior to attack to proactively resist adversarial perturbations. Unfortunately, their practical applicability remains questionable due to several limitations, including (1) reliance on well-trained classifiers as surrogates to provide robustness priors, (2) substantial computational overhead arising from iterative optimization or trained generators for robustification, and (3) limited interpretability of the optimization- or generation-based robustification processes. Inspired by recent studies revealing a positive correlation between texture intensity and the robustness of benign samples, we show that image sharpening alone can efficiently robustify images. To the best of our knowledge, this is the first surrogate-free, optimization-free, generator-free, and human-interpretable robustification approach. Extensive experiments demonstrate that sharpening yields remarkable robustness gains with low computational cost, especially in transfer scenarios.
format Preprint
id arxiv_https___arxiv_org_abs_2603_25244
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Efficient Preemptive Robustification with Image Sharpening
Liang, Jiaming
Pun, Chi-Man
Computer Vision and Pattern Recognition
Despite their great success, deep neural networks rely on high-dimensional, non-robust representations, making them vulnerable to imperceptible perturbations, even in transfer scenarios. To address this, both training-time defenses (e.g., adversarial training and robust architecture design) and post-attack defenses (e.g., input purification and adversarial detection) have been extensively studied. Recently, a limited body of work has preliminarily explored a pre-attack defense paradigm, termed preemptive robustification, which introduces subtle modifications to benign samples prior to attack to proactively resist adversarial perturbations. Unfortunately, their practical applicability remains questionable due to several limitations, including (1) reliance on well-trained classifiers as surrogates to provide robustness priors, (2) substantial computational overhead arising from iterative optimization or trained generators for robustification, and (3) limited interpretability of the optimization- or generation-based robustification processes. Inspired by recent studies revealing a positive correlation between texture intensity and the robustness of benign samples, we show that image sharpening alone can efficiently robustify images. To the best of our knowledge, this is the first surrogate-free, optimization-free, generator-free, and human-interpretable robustification approach. Extensive experiments demonstrate that sharpening yields remarkable robustness gains with low computational cost, especially in transfer scenarios.
title Efficient Preemptive Robustification with Image Sharpening
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2603.25244