LaserGuider: A Laser Based Physical Backdoor Attack against Deep Neural Networks

TL;DR

LaserGuider introduces a novel laser-based physical backdoor attack on DNNs, enabling remote, stealthy, and flexible manipulation of models, demonstrated effectively on traffic sign recognition systems.

Contribution

This work presents LaserGuider, a new laser-based backdoor attack with optimized parameters, and introduces LaserMark, a dataset of traffic signs with laser spots for research.

Findings

Achieves over 90% attack success rate on traffic sign DNNs.

Demonstrates high temporal stealthiness and mobility of laser triggers.

Provides a new dataset for backdoor attack research.

Abstract

Backdoor attacks embed hidden associations between triggers and targets in deep neural networks (DNNs), causing them to predict the target when a trigger is present while maintaining normal behavior otherwise. Physical backdoor attacks, which use physical objects as triggers, are feasible but lack remote control, temporal stealthiness, flexibility, and mobility. To overcome these limitations, in this work, we propose a new type of backdoor triggers utilizing lasers that feature long-distance transmission and instant-imaging properties. Based on the laser-based backdoor triggers, we present a physical backdoor attack, called LaserGuider, which possesses remote control ability and achieves high temporal stealthiness, flexibility, and mobility. We also introduce a systematic approach to optimize laser parameters for improving attack effectiveness. Our evaluation on traffic sign recognition DNNs, critical in autonomous vehicles, demonstrates that LaserGuider with three different laser-based triggers achieves over 90% attack success rate with negligible impact on normal inputs. Additionally, we release LaserMark, the first dataset of real world traffic signs stamped with physical laser spots, to support further research in backdoor attacks and defenses.