面向安全感知端到端驾驶的可靠策略迁移深度强化学习
Reliable Policy Transfer for Safety-Aware End-to-End Driving with Deep Reinforcement Learning
把逐边异方差不确定性与 critic 集成的不确定性信号统一接入关系图注意力、策略熵门控与迁移对齐，为视觉端到端驾驶提供了一种可复用的「控制层可靠性接口」设计；做闭环驾驶策略学习时可直接借鉴其不确定性驱动探索与少样本域迁移的做法。
原文依据 the same uncertainty signal informs relational attention, gates policy entropy, and regularizes transfer alignment
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论面向自动驾驶的端到端（E2E）强化学习（RL）在分布偏移下仍难以兼顾安全性与泛化能力，原因在于感知负担过重的编码器、稀疏奖励以及临时性的不确定性处理会导致闭环行为脆弱。本文提出一个统一的深度强化学习（DRL）框架，其核心是控制层可靠性接口：同一不确定性信号同时用于关系注意力、策略熵门控与迁移对齐正则化。以自车为中心的关系图通过不确定性加权注意力对运动学、车道几何与语义信息进行编码，刻画各智能体的影响，并输出紧凑的控制状态。多目标可微奖励结合不确定性项来塑造安全性、行进效率与舒适性。由逐边异方差方差与 critic 集成分别捕获的偶然不确定性与认知不确定性，用于调制策略熵以实现风险感知探索。因果-语义迁移目标在跨域之间对齐动作、注意力与不确定性统计量，并借助元学习初始化实现少样本自适应。在多种城镇、交通与天气条件下的闭环城市驾驶中，该框架相较强基线提升了成功率、减少了违规，并在获得更高冲突时间的同时降低了横向偏差。
英文原文摘要 1313 字符
End-to-End (E2E) Reinforcement Learning (RL) for autonomous driving still struggles with safety and generalization under distribution shift, as perception-heavy encoders, sparse rewards, and ad hoc uncertainty handling yield brittle closed-loop behavior. This work introduces a unified Deep RL (DRL) framework built around a control-layer reliability interface where the same uncertainty signal informs relational attention, gates policy entropy, and regularizes transfer alignment. An ego-centric relational graph encodes agent influence via uncertainty-weighted attention over kinematics, lane geometry, and semantics, producing a compact control state. A multi-objective differentiable reward shapes safety, progress, and comfort with an uncertainty term. Aleatoric and epistemic uncertainty, captured through per-edge heteroscedastic variance and a critic ensemble, modulate policy entropy for risk-aware exploration. A causal-semantic transfer objective aligns actions, attention, and uncertainty statistics across domains with meta-learned initialization for few-shot adaptation. In closed-loop urban driving across varied towns, traffic, and weather, the framework improves success rate, reduces infractions, and achieves higher time-to-conflict combined with lower lateral deviation over strong baselines.