OptiMVMap：基于最优多车视角的离线向量化地图构建
OptiMVMap: Offline Vectorized Map Construction via Optimal Multi-vehicle Perspectives
将多车视角融合建模为「先选择后融合」问题，其不确定性引导的辅助车选择与跨车注意力对齐机制，可直接用于BEV地图构建中缓解遮挡与位姿噪声。
原文依据 OptiMVMap, which reformulates multi-vehicle mapping as a select-then-fuse problem
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论离线向量化地图是高精度自动驾驶与地图服务的关键基础设施。现有方法主要依赖单自车轨迹，从根本上受限于视角不足：基于记忆的方法虽通过聚合自车轨迹帧延长了观测时间，却缺乏揭示遮挡区域所需的空间多样性。引入周围车辆的视角可提供互补信息，但朴素融合会带来三大挑战：大候选池导致的计算开销、近共线视角造成的冗余，以及位姿误差与遮挡伪影引入的噪声。本文提出OptiMVMap，将多车地图构建重构为“先选择后融合”问题以系统性地应对上述挑战。最优车辆选择（OVS）模块策略性地识别一个紧凑的辅助车集合，该集合能最大程度降低自车在遮挡区域的不确定性，从而解决计算与冗余挑战。跨车注意力（CVA）与语义感知噪声滤波器（SNF）则在BEV级融合前执行位姿容忍的对齐与伪影抑制，以解决噪声挑战。该针对性流程以远少于无差别聚合的视角数量，生成更完整且拓扑更忠实的地图。在nuScenes与Argoverse2上，OptiMVMap分别将MapTRv2提升+10.5平均精度均值与+9.3平均精度均值，并在nuScenes上以+6.2平均精度均值与+3.8平均精度均值超越记忆增强基线MVMap与HRMapNet。这些结果表明，不确定性引导的辅助车选择对于高效准确的多车向量化地图构建至关重要。代码已发布于https://github.com/DanZeDong/OptiMVMap。
英文原文摘要 1766 字符
Offline vectorized maps constitute critical infrastructure for high-precision autonomous driving and mapping services. Existing approaches rely predominantly on single ego-vehicle trajectories, which fundamentally suffer from viewpoint insufficiency: while memory-based methods extend observation time by aggregating ego-trajectory frames, they lack the spatial diversity needed to reveal occluded regions. Incorporating views from surrounding vehicles offers complementary perspectives, yet naive fusion introduces three key challenges: computational cost from large candidate pools, redundancy from near-collinear viewpoints, and noise from pose errors and occlusion artifacts. We present OptiMVMap, which reformulates multi-vehicle mapping as a select-then-fuse problem to address these challenges systematically. An Optimal Vehicle Selection (OVS) module strategically identifies a compact subset of helpers that maximally reduce ego-centric uncertainty in occluded regions, addressing computation and redundancy challenges. Cross-Vehicle Attention (CVA) and Semantic-aware Noise Filter (SNF) then perform pose-tolerant alignment and artifact suppression before BEV-level fusion, addressing the noise challenge. This targeted pipeline yields more complete and topologically faithful maps with substantially fewer views than indiscriminate aggregation. On nuScenes and Argoverse2, OptiMVMap improves MapTRv2 by +10.5 mAP and +9.3 mAP, respectively, and surpasses memory-augmented baselines MVMap and HRMapNet by +6.2 mAP and +3.8 mAP on nuScenes. These results demonstrate that uncertainty-guided selection of helper vehicles is essential for efficient and accurate multi-vehicle vectorized mapping. The code is released at https://github.com/DanZeDong/OptiMVMap.