Research map / Label-efficient and robust learning / Recognition and long-tail
ADMap: Anti-disturbance Framework for Vectorized HD Map Construction
Haotian Hu, Fanyi Wang, Yaonong Wang, Laifeng Hu, Jingwei Xu, Zhiwang Zhang
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 all three modules (IIA, MPN, VDDL) to MapTR (ablation) | MapTR camera+lidar · nuScenes | mAP: 62.5 -> 68.0 | Improved |
| add instance interactive attention (IIA) (ablation) | MapTR · nuScenes | mAP: 62.5 -> 65.4 | Improved |
| add vector direction difference loss (VDDL) (ablation) | MapTR · nuScenes | mAP: 62.5 -> 64.6 | Improved |
| use IIA instead of decoupled self-attention (DSA) (ablation) | MapTR · nuScenes | mAP: 64.1 -> 65.4 | Improved |
| use MPN instead of SECOND neck (ablation) | MapTR · nuScenes | mAP: 63.7 -> 64.5 | Improved |
| increase VDDL weight from 1.0 to 2.0 or more (ablation) | ADMap camera-only · nuScenes | mAP: 53.3 -> 51.7 | Hurt |
| increase MPN downsampling layers from 2 to 3 (ablation) | ADMapv2 · nuScenes | mAP: 70.3 -> 70.5 mAP, FPS 5.8 -> 5.3 | Mixed |
| apply ADMap framework to MapTR/MapTRv2 | MapTR, MapTRv2 · nuScenes camera-only | mAP: 50.3 -> 54.5; 61.5 -> 62.9 | Improved |
| apply ADMap framework | MapTR, MapTRv2 · nuScenes jitter | AJP: 6.76 -> 4.08; 4.53 -> 2.44 | Improved |
| apply ADMap framework to MapTR | MapTR · nuScenes camera-only | FPS: 15.1 -> 14.8 | Hurt |
Closest papers by meaning
- Leveraging SD Map to Augment HD Map-based Trajectory Prediction (CVPR 2025 · similarity 0.90)
- MapExpert: Online HD Map Construction with Simple and Efficient Sparse Map Element Expert (AAAI 2025 · similarity 0.90)
- InteractionMap: Improving Online Vectorized HDMap Construction with Interaction (CVPR 2025 · similarity 0.88)
- Producing and Leveraging Online Map Uncertainty in Trajectory Prediction (CVPR 2024 · similarity 0.88)
- Uncertainty-Instructed Structure Injection for Generalizable HD Map Construction (CVPR 2025 · similarity 0.87)
- SDTagNet: Leveraging Text-Annotated Navigation Maps for Online HD Map Construction (NeurIPS 2025 · similarity 0.87)
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.
