MetAdv: A Unified and Interactive Adversarial Testing Platform for Autonomous Driving

2025-08-03

AI Summary

Evaluating the robustness of autonomous driving systems under adversarial conditions remains challenging. This paper proposes a cyber-physical end-to-end adversarial testing platform featuring a three-layer closed-loop architecture—simulation, hardware-in-the-loop, and human-factor integration. For the first time, it incorporates driver physiological feedback (e.g., EEG and eye-tracking signals) into the adversarial evolution testing pipeline, enabling dynamic coupling among synthetic scenario generation, real-vehicle response, and human cognitive state estimation. The platform is compatible with mainstream autonomous driving stacks—including Apollo and Tesla—and integrates high-fidelity 3D scene modeling, naturalistic driving behavior capture, and multimodal perception validation within a unified closed-loop framework. Experimental results demonstrate that the method significantly improves quantitative robustness assessment under adversarial conditions and enhances the interpretability of human–autonomy trust evaluation.

Technology Category

• Multiagent Systems: Adversarial Agents
• Computer Vision: Adversarial Attacks & Robustness
• Machine Learning: Adversarial Learning & Robustness

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Security and Privacy: Large-scale security measurements

Abstract

Evaluating and ensuring the adversarial robustness of autonomous driving (AD) systems is a critical and unresolved challenge. This paper introduces MetAdv, a novel adversarial testing platform that enables realistic, dynamic, and interactive evaluation by tightly integrating virtual simulation with physical vehicle feedback. At its core, MetAdv establishes a hybrid virtual-physical sandbox, within which we design a three-layer closed-loop testing environment with dynamic adversarial test evolution. This architecture facilitates end-to-end adversarial evaluation, ranging from high-level unified adversarial generation, through mid-level simulation-based interaction, to low-level execution on physical vehicles. Additionally, MetAdv supports a broad spectrum of AD tasks, algorithmic paradigms (e.g., modular deep learning pipelines, end-to-end learning, vision-language models). It supports flexible 3D vehicle modeling and seamless transitions between simulated and physical environments, with built-in compatibility for commercial platforms such as Apollo and Tesla. A key feature of MetAdv is its human-in-the-loop capability: besides flexible environmental configuration for more customized evaluation, it enables real-time capture of physiological signals and behavioral feedback from drivers, offering new insights into human-machine trust under adversarial conditions. We believe MetAdv can offer a scalable and unified framework for adversarial assessment, paving the way for safer AD.

Problem

• Evaluating adversarial robustness in autonomous driving systems
• Integrating virtual simulation with physical vehicle feedback
• Assessing human-machine trust under adversarial conditions

Innovation

• Hybrid virtual-physical sandbox for testing
• Three-layer closed-loop adversarial evaluation
• Human-in-the-loop with physiological feedback