OBJVanish: Prompt-Driven Generation of Physically Realizable 3D LiDAR-Invisible Objects

LiDAR-based 3D object detectors are fundamental to autonomous driving, where missed detections pose severe safety risks. While adversarial attacks are crucial for evaluating the robustness of these detectors, existing point-level perturbation methods rarely cause complete object disappearance and prove difficult to implement in physical environments. We introduce OBJVanish, a prompt-driven text-to-3D adversarial generation framework that enables physically realizable attacks by generating 3D object models that are effectively invisible to LiDAR-based 3D object detectors. We first conduct a systematic empirical study of detection vulnerability in LiDAR-based 3D object detectors, revealing multi-object compositions as the dominant factor. Based on this analysis, the proposed framework iteratively refines text prompts—optimizing verbs, objects, and poses—to generate LiDAR-invisible pedestrian instances as representative vulnerable road users under physical constraints. To ensure realizability, the framework operates over a curated pool of representative real-world 3D object models and restricts generation to their valid combinations. Extensive experiments show that OBJVanish consistently evades six state-of-the-art (SOTA) LiDAR-based 3D object detectors in both simulation and real-world physical settings, exposing critical vulnerabilities in safety-critical detection systems.

Laser sensors (LiDAR) are the "eyes" of self-driving cars, crucial for detecting pedestrians and ensuring road safety. However, current testing methods struggle to simulate how everyday objects might accidentally confuse these systems, leaving dangerous security gaps. In this study, we developed a testing framework called OBJVanish that uses generative AI to find these blind spots. We discovered a critical flaw: when a pedestrian carries or interacts with common items—like holding an umbrella or pushing a wheelchair—the combined shape disrupts the AI’s recognition, causing the person to completely "vanish" from the vehicle’s radar. We successfully tested these set-ups in both digital simulations and real-world driving environments, proving that ordinary pedestrians with specific everyday items can bypass top-tier autonomous sensors. This research exposes vital vulnerabilities, helping engineers build much safer and more reliable self-driving systems.