[Submitted on 10 Nov 2025 (v1), last revised 29 Jul 2026 (this version, v3)]

View PDF HTML (experimental)

Abstract:Clean-image backdoor attacks, which use only label manipulation in training datasets to compromise deep neural networks, pose a significant threat to security-critical applications. A critical flaw in existing methods is that the poison rate required for a successful attack induces a proportional, and thus noticeable, drop in Clean Accuracy (CA), undermining their stealthiness. This paper presents a new paradigm for clean-image attacks that minimizes this accuracy degradation by optimizing the trigger itself. We introduce Generative Clean-Image Backdoors (GCB), a framework that uses a conditional InfoGAN to identify naturally occurring image features that can serve as potent and stealthy triggers. By ensuring these triggers are easily separable from benign task-related features, GCB enables a victim model to learn the backdoor from an extremely small set of poisoned examples, resulting in a CA drop of less than 1%. Our experiments demonstrate GCB's remarkable versatility, successfully adapting to six datasets, five architectures, and four tasks, including the first demonstration of clean-image backdoors in regression and segmentation. GCB also exhibits resilience against most of the existing backdoor defenses.

Submission history

From: Binyan Xu [view email]
[v1] Mon, 10 Nov 2025 15:37:44 UTC (6,081 KB)
[v2] Tue, 11 Nov 2025 06:02:02 UTC (6,081 KB)
[v3] Wed, 29 Jul 2026 07:52:20 UTC (6,089 KB)