Research map / 3D vision / 3D generation
OptiMVMap: Offline Vectorized Map Construction via Optimal Multi-vehicle Perspectives
Zedong Dan, Zijie Wang, Wei Zhang, Xiangru Lin, Weiming Zhang, Xiao Tan, Jingdong Wang, Liang Lin, Guanbin Li
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add naive fusion of ego and non-ego features (ablation) | MapTRv2 · nuScenes (1/4 subset) | mAP: +1.8 mAP | Improved |
| add cross-vehicle attention (CVA) (ablation) | MapTRv2 + naive fusion · nuScenes (1/4 subset) | mAP: 39.5 -> 42.1 | Improved |
| add semantic-aware noise filter (SNF) (ablation) | MapTRv2 + CVA · nuScenes (1/4 subset) | mAP: 42.1 -> 44.1 | Improved |
| replace BEV sum pooling with depth-semantic decoupled max pooling (ablation) | MapTRv2 + CVA + SNF · nuScenes (1/4 subset) | mAP: 44.1 -> 45.0 | Improved |
| add optimal vehicle selection (OVS) (ablation) | MapTRv2 + CVA + SNF + pooling · nuScenes (1/4 subset) | mAP: 45.0 -> 49.8 | Improved |
| select helpers with OVS instead of random (ablation) | MapTRv2 + OptiMVMap · nuScenes | mAP: +6.9 mAP over Random | Improved |
| select helpers with OVS instead of closest (ablation) | MapTRv2 + OptiMVMap · nuScenes | mAP: +4.4 mAP over Closest | Improved |
| increase helper count K from 1 to 2 (ablation) | MapTRv2 + OptiMVMap · nuScenes | mAP: 66.2 -> 71.0 | Improved |
| increase helper count K from 5 to 7 (ablation) | MapTRv2 + OptiMVMap · nuScenes | mAP: 73.9 -> 74.3 | No clear effect |
| add query initialization (QI) from MGMap (ablation) | MapTRv2 + OptiMVMap · nuScenes | mAP: 71.0 -> 72.0 | Improved |
| add OptiMVMap plug-in (ablation) | VectorMapNet · nuScenes | mAP: 40.9 -> 55.1 | Improved |
| add OptiMVMap plug-in (ablation) | MapTRv2 · Argoverse 2 | mAP: 64.3 -> 73.6 | Improved |
Closest papers by meaning
- Uncertainty-Instructed Structure Injection for Generalizable HD Map Construction (CVPR 2025 · similarity 0.87)
- Map It Anywhere: Empowering BEV Map Prediction using Large-scale Public Datasets (NeurIPS 2024 · similarity 0.87)
- MapExpert: Online HD Map Construction with Simple and Efficient Sparse Map Element Expert (AAAI 2025 · similarity 0.86)
- SafeMap: Robust HD Map Construction from Incomplete Observations (ICML 2025 · similarity 0.86)
- Producing and Leveraging Online Map Uncertainty in Trajectory Prediction (CVPR 2024 · similarity 0.86)
- InteractionMap: Improving Online Vectorized HDMap Construction with Interaction (CVPR 2025 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
