IU: Imperceptible Universal Backdoor Attack
Hsin Lin, Yan-Lun Chen, Ren-Hung Hwang, Chia-Mu Yu

TL;DR
This paper presents a novel, imperceptible universal backdoor attack on deep neural networks that uses graph convolutional networks to generate stealthy, effective perturbations across multiple classes, demonstrating high success rates with minimal poisoning.
Contribution
Introduces a GCN-based framework for creating invisible, scalable universal backdoors that balance stealth and attack success, advancing backdoor attack techniques.
Findings
Achieves up to 91.3% attack success rate on ImageNet-1K
Operates with as low as 0.16% poisoning rate
Evades current state-of-the-art defenses
Abstract
Backdoor attacks pose a critical threat to the security of deep neural networks, yet existing efforts on universal backdoors often rely on visually salient patterns, making them easier to detect and less practical at scale. In this work, we introduce a novel imperceptible universal backdoor attack that simultaneously controls all target classes with minimal poisoning while preserving stealth. Our key idea is to leverage graph convolutional networks (GCNs) to model inter-class relationships and generate class-specific perturbations that are both effective and visually invisible. The proposed framework optimizes a dual-objective loss that balances stealthiness (measured by perceptual similarity metrics such as PSNR) and attack success rate (ASR), enabling scalable, multi-target backdoor injection. Extensive experiments on ImageNet-1K with ResNet architectures demonstrate that our method…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsAdversarial Robustness in Machine Learning · Digital Media Forensic Detection · Advanced Neural Network Applications
