跳到正文

STA: A Split Temporal Adversary Method for Evaluating Robustness of Multimodal RL-based Driving Policies

机器人 0 次浏览

原文:STA: A Split Temporal Adversary Method for Evaluating Robustness of Multimodal RL-based Driving Policies

作者:Shenghao Xiong, Xiaolin Chang

来源:arXiv cs.RO(机器人)

正文

Computer Science > Cryptography and Security

arXiv:2611.20765v1 (cs)

[Submitted on 16 Nov 2026]

Title:STA: A Split Temporal Adversary Method for Evaluating Robustness of Multimodal RL-based Driving Policies

Authors:Shenghao Xiong, Xiaolin Chang

View a PDF of the paper titled STA: A Split Temporal Adversary Method for Evaluating Robustness of Multimodal RL-based Driving Policies, by Shenghao Xiong and 1 other authors

View PDF

HTML (experimental)

Abstract:Multimodal reinforcement learning (MRL) offers a promising paradigm for autonomous driving by jointly leveraging complementary sensor observations to learn effective decision-making policies. However, its effectiveness critically depends on temporal alignment across modalities, making temporal inconsistency a fundamental vulnerability. Even subtle misalignment or latency can corrupt fused representations and ultimately lead to decision failures. Existing studies have explored adversarial attacks to induce cross-modal temporal inconsistencies, thereby enabling systematic evaluation of the temporal robustness of MRL policies. However, existing methods remain limited in generality, realism, and adaptivity. To address these limitations, we propose the Split Temporal Adversary (STA), which formulates multimodal temporal attacks as a unified decision process that jointly optimizes attack timing and modality selection under realistic policy-aware and budget-constrained settings. STA includes two mechanisms: a) a unified multimodal temporal perturbation mechanism that delays observations via historical states, removing the need for modality-specific noise; b) a split multimodal adversarial decision mechanism that jointly optimizes attack timing and modality selection through adaptive optimization while leveraging the victim policy’s clean actions to identify sensitive decision patterns. We provide a theoretical analysis of the induced distribution shift and a modality selection principle. Experiments in the CARLA simulator show that STA more than doubles the collision rate of the strongest baseline against PPO while using 17.8% fewer attack attempts, and that it generalizes to a VLM-based driving policy. These results reveal temporal consistency as a critical vulnerability, highlighting temporal robustness as essential for safe autonomous driving.

Comments:

10 pages, 6 figures, submitted for publication

Subjects:

Cryptography and Security (cs.CR); Machine Learning (cs.LG); Robotics (cs.RO)

Cite as:

arXiv:2611.20765 [cs.CR]

(or

arXiv:2611.20765v1 [cs.CR] for this version)

https://doi.org/10.48550/arXiv.2611.20765

Focus to learn more

arXiv-issued DOI via DataCite (pending registration)

主题

安全 · 机器学习 · 机器人


由「前沿雷达」于 2026-11-16 采集。正文取自原文页面,已保留出处链接。标题与正文版权归原作者所有。

评论

加载中…