In-the-Wild Camouflage Attack on Vehicle Detectors through Controllable Image Editing

TL;DR

This paper introduces a controllable image-editing framework for vehicle camouflage attacks that significantly reduces detector accuracy while maintaining stealthiness, demonstrating strong effectiveness and transferability in real-world scenarios.

Contribution

It presents a novel conditional image-editing approach using ControlNet for vehicle camouflage attacks, improving attack success and stealthiness over prior methods.

Findings

Achieves over 38% AP50 decrease in vehicle detection accuracy.

Effectively generalizes to unseen black-box detectors.

Shows promising transferability to physical-world scenarios.

Abstract

Deep neural networks (DNNs) have achieved remarkable success in computer vision but remain highly vulnerable to adversarial attacks. Among them, camouflage attacks manipulate an object's visible appearance to deceive detectors while remaining stealthy to humans. In this paper, we propose a new framework that formulates vehicle camouflage attacks as a conditional image-editing problem. Specifically, we explore both image-level and scene-level camouflage generation strategies, and fine-tune a ControlNet to synthesize camouflaged vehicles directly on real images. We design a unified objective that jointly enforces vehicle structural fidelity, style consistency, and adversarial effectiveness. Extensive experiments on the COCO and LINZ datasets show that our method achieves significantly stronger attack effectiveness, leading to more than 38% AP50 decrease, while better preserving vehicle structure and improving human-perceived stealthiness compared to existing approaches. Furthermore, our framework generalizes effectively to unseen black-box detectors and exhibits promising transferability to the physical world. Project page is available at https://humansensinglab.github.io/CtrlCamo