Breaking the Stealth-Potency Trade-off in Clean-Image Backdoors with Generative Trigger Optimization
This addresses a critical security flaw in deep neural networks for applications where stealth is essential, representing a new paradigm rather than an incremental improvement.
The paper tackles the trade-off between stealth and effectiveness in clean-image backdoor attacks by introducing a generative framework that optimizes triggers to minimize accuracy degradation, achieving a drop in clean accuracy of less than 1% while adapting to multiple datasets, architectures, and tasks.
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.