Transferable Dual-Domain Feature Importance Attack against AI-Generated Image Detector

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Zhu, Weiheng, Cao, Gang, Liu, Jing, Yu, Lifang, Weng, Shaowei
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914164457340928
author Zhu, Weiheng
Cao, Gang
Liu, Jing
Yu, Lifang
Weng, Shaowei
author_facet Zhu, Weiheng
Cao, Gang
Liu, Jing
Yu, Lifang
Weng, Shaowei
contents Recent AI-generated image (AIGI) detectors achieve impressive accuracy under clean condition. In view of antiforensics, it is significant to develop advanced adversarial attacks for evaluating the security of such detectors, which remains unexplored sufficiently. This letter proposes a Dual-domain Feature Importance Attack (DuFIA) scheme to invalidate AIGI detectors to some extent. Forensically important features are captured by the spatially interpolated gradient and frequency-aware perturbation. The adversarial transferability is enhanced by jointly modeling spatial and frequency-domain feature importances, which are fused to guide the optimization-based adversarial example generation. Extensive experiments across various AIGI detectors verify the cross-model transferability, transparency and robustness of DuFIA.
format Preprint
id arxiv_https___arxiv_org_abs_2511_15571
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Transferable Dual-Domain Feature Importance Attack against AI-Generated Image Detector
Zhu, Weiheng
Cao, Gang
Liu, Jing
Yu, Lifang
Weng, Shaowei
Computer Vision and Pattern Recognition
Cryptography and Security
Recent AI-generated image (AIGI) detectors achieve impressive accuracy under clean condition. In view of antiforensics, it is significant to develop advanced adversarial attacks for evaluating the security of such detectors, which remains unexplored sufficiently. This letter proposes a Dual-domain Feature Importance Attack (DuFIA) scheme to invalidate AIGI detectors to some extent. Forensically important features are captured by the spatially interpolated gradient and frequency-aware perturbation. The adversarial transferability is enhanced by jointly modeling spatial and frequency-domain feature importances, which are fused to guide the optimization-based adversarial example generation. Extensive experiments across various AIGI detectors verify the cross-model transferability, transparency and robustness of DuFIA.
title Transferable Dual-Domain Feature Importance Attack against AI-Generated Image Detector
topic Computer Vision and Pattern Recognition
Cryptography and Security
url https://arxiv.org/abs/2511.15571