用学习到的驾驶技能生成更真实的对抗性交通场景，提升自动驾驶安全性。
SEAL: Towards Safe Autonomous Driving via Skill-Enabled Adversary Learning for Closed-Loop Scenario Generation
- 基于人类驾驶技能和学习目标函数生成对抗性场景
- 在真实与分布外场景中使主车任务成功率提升超20%
- 适合自动驾驶安全验证与强化学习研究者使用
自动驾驶系统及其组件的验证与测试日益重要，随着其在现实世界中的普及。安全关键场景生成是通过闭环训练增强自动驾驶策略鲁棒性的关键方法。然而，现有方法依赖于简化的目标函数，导致生成的对抗行为过于激进或缺乏反应性。为此，我们提出SEAL，一种基于学习目标函数和类人对抗技能的场景扰动方法。SEAL生成的场景比当前最优基线更真实，在真实世界、分布内及分布外场景中，主车任务成功率提升超过20%。为促进后续研究，我们开源了代码与工具：https://github.com/cmubig/SEAL。
原文摘要 · Abstract (English)
Verification and validation of autonomous driving (AD) systems and components is of increasing importance, as such technology increases in real-world prevalence. Safety-critical scenario generation is a key approach to robustify AD policies through closed-loop training. However, existing approaches for scenario generation rely on simplistic objectives, resulting in overly-aggressive or non-reactive adversarial behaviors. To generate diverse adversarial yet realistic scenarios, we propose SEAL, a scenario perturbation approach which leverages learned objective functions and adversarial, human-like skills. SEAL-perturbed scenarios are more realistic than SOTA baselines, leading to improved ego task success across real-world, in-distribution, and out-of-distribution scenarios, of more than 20%. To facilitate future research, we release our code and tools: https://github.com/cmubig/SEAL
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。