English(EN) MTA-RL: Robust Urban Driving via Multi-modal Transformer-based 3D Affordances and Reinforcement Learning

MTA-RL框架通过多模态AI增强城市驾驶

作者 PulseAugur 编辑部 · [2 个来源] · 2026-05-11 08:28

研究人员开发了MTA-RL，一个新颖的框架，它集成了多模态Transformer-based 3D可供性与强化学习，以实现鲁棒的城市自动驾驶。该方法融合了RGB图像和LiDAR数据，以预测明确的、几何感知的可供性，为强化学习策略创建了一个结构化的观测空间。在CARLA模拟器中的评估表明，与现有基线相比，MTA-RL在样本效率、稳定性和零样本泛化方面表现更优。 AI

影响引入了一种连接自动驾驶感知与控制的新颖方法，提高了样本效率和泛化能力。

排序理由该集群包含一篇详细介绍自动驾驶新AI框架的学术论文。

在 arXiv cs.AI 阅读 →

AI 生成摘要 · Google Gemini · 来自 2 个来源。我们如何撰写摘要 →

报道来源 [2]

arXiv cs.AI TIER_1 English(EN) · Ostap Okhrin · 2026-05-11 08:28

MTA-RL: Robust Urban Driving via Multi-modal Transformer-based 3D Affordances and Reinforcement Learning

Robust urban autonomous driving requires reliable 3D scene understanding and stable decision-making under dense interactions. However, existing end-to-end models lack interpretability, while modular pipelines suffer from error propagation across brittle interfaces. This paper pro…
Hugging Face Daily Papers TIER_1 English(EN) · 2026-05-11 08:28

MTA-RL: Robust Urban Driving via Multi-modal Transformer-based 3D Affordances and Reinforcement Learning

Robust urban autonomous driving requires reliable 3D scene understanding and stable decision-making under dense interactions. However, existing end-to-end models lack interpretability, while modular pipelines suffer from error propagation across brittle interfaces. This paper pro…

报道来源 [2]

MTA-RL: Robust Urban Driving via Multi-modal Transformer-based 3D Affordances and Reinforcement Learning

MTA-RL: Robust Urban Driving via Multi-modal Transformer-based 3D Affordances and Reinforcement Learning

相关实体

相关话题