视觉与机器人

自动驾驶

自动驾驶感知、规划、BEV、占用预测、激光雷达和仿真评测。

2026-05-28 至 2026-05-28 共收录 2 篇信号源：cs.RO, cs.CV, eess.IV, cs.AI

1. 仿真评测 2 篇

2604.05378 2026-05-28 cs.CL cs.CV 79%

ICR-Drive: Instruction Counterfactual Robustness for End-to-End Language-Driven Autonomous Driving

ICR-Drive：面向端到端语言驱动自动驾驶的指令反事实鲁棒性

Kaiser Hamid, Can Cui, Nade Liang

机构 * Texas Tech University（德克萨斯科技大学）； Bosch Center for Artificial Intelligence (BCAI)（博世人工智能中心（BCAI））

专题命中仿真评测：autonomous driving(title,abstract)；分类 cs.CV

AI总结提出ICR-Drive框架，通过生成四类扰动指令（改写、歧义、噪声、误导）并基于CARLA仿真评估，揭示语言条件驾驶模型对指令变化的脆弱性。

Journal ref Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, 2026, pp. 872-880

2503.04863 2026-05-28 cs.CV cs.AI 62%

Manboformer: Learning Gaussian Representations via Spatial-temporal Attention Mechanism

Manboformer: 通过时空注意力机制学习高斯表示

Ziyue Zhao, Qining Qi, Jianfa Ma

专题命中仿真评测：autonomous driving(abstract)；分类 cs.CV、cs.AI

AI总结针对自动驾驶3D语义占用预测中高斯表示性能不足的问题，提出利用时空自注意力机制优化GaussianFormer，以提升模型性能。

Comments After careful self-check, we found several unnoticed deficiencies and incomplete discussions in this manuscript. To ensure the rigor and accuracy of academic results, we decide to withdraw this preprint. A refined, complete, and rigorous version will be submitted soon