OptiMVMap: Offline Vectorized Map Construction via Optimal Multi-vehicle Perspectives

2026-04-18

AI Summary

This work addresses the limitations of single-view vehicle trajectory–based vectorized map construction, which suffers from occlusion, and multi-vehicle fusion approaches, which often incur computational redundancy, overlapping viewpoints, and pose noise. To overcome these challenges, the authors propose a “select-and-fuse” framework featuring an uncertainty-guided Optimal Vehicle Selection (OVS) mechanism that substantially reduces the number of required viewpoints. Robust alignment and efficient fusion in bird’s-eye-view (BEV) space are achieved through Cross-Vehicle Attention (CVA) and a Semantic-aware Noise Filter (SNF). Evaluated on nuScenes and Argoverse2, the method improves MapTRv2 by +10.5 mAP and +9.3 mAP, respectively, outperforming memory-augmented baselines such as MVMap and HRMapNet, and significantly enhancing both map completeness and topological accuracy.

Technology Category

• Computer Vision: Multi-modal Vision
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Planning, Routing, and Scheduling: Model-Based Reasoning

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

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.

Problem

• vectorized map
• multi-vehicle perception
• viewpoint insufficiency
• occlusion
• map construction

Innovation

• Optimal Vehicle Selection
• Cross-Vehicle Attention
• Semantic-aware Noise Filter
• Vectorized Map Construction
• Multi-vehicle Perspective Fusion