CISPA
Browse

IAP: Invisible Adversarial Patch Attack through Perceptibility-Aware Localization and Perturbation Optimization

Download (11.73 MB)
conference contribution
posted on 2025-10-07, 09:27 authored by Subrat Kishore Dutta, Xiao ZhangXiao Zhang
Despite modifying only a small localized input region, adversarial patches can drastically change the prediction of computer vision models. However, prior methods either cannot perform satisfactorily under targeted attack scenarios or fail to produce contextually coherent adversarial patches, causing them to be easily noticeable by human examiners and insufficiently stealthy against automatic patch defenses. In this paper, we introduce IAP, a novel attack framework that generates highly invisible adversarial patches based on perceptibility-aware localization and perturbation optimization schemes. Specifically, IAP first searches for a proper location to place the patch by leveraging classwise localization and sensitivity maps, balancing the susceptibility of patch location to both victim model prediction and human visual system, then employs a perceptibility-regularized adversarial loss and a gradient update rule that prioritizes color constancy for optimizing invisible perturbations. Comprehensive experiments across various image benchmarks and model architectures demonstrate that IAP consistently achieves competitive attack success rates in targeted settings with significantly improved patch invisibility compared to existing baselines. In addition to being highly imperceptible to humans, IAP is shown to be stealthy enough to render several state-of-the-art patch defenses ineffective.

History

Primary Research Area

  • Trustworthy Information Processing

Name of Conference

IEEE International Conference on Computer Vision (ICCV)

CISPA Affiliation

  • Yes

BibTeX

@conference{Dutta:Zhang:2025, title = "IAP: Invisible Adversarial Patch Attack through Perceptibility-Aware Localization and Perturbation Optimization", author = "Dutta, Subrat Kishore" AND "Zhang, Xiao", year = 2025, month = 10 }

Usage metrics

    Categories

    No categories selected

    Licence

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC