Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

MGMap: Mask-Guided Learning for Online Vectorized HD Map Construction

Xiaolu Liu, Song Wang, Wentong Li, Ruizi Yang, Junbo Chen, Jianke Zhu

TL;DR

MGMap tackles the challenge of precise online HD map vectorization under subtle and sparse annotations by introducing mask-guided learning. It combines an Enhanced Multi-Level BEV feature extractor with a Mask-Activated Instance (MAI) decoder and a Position-Guided Mask Patch Refinement (PG-MPR) module to perform coarse instance-level and fine-grained point-level localization. The method jointly leverages learned instance masks and binary mask features to highlight informative regions and refine point locations through ROI-based patch features, achieving substantial mAP gains across nuScenes and Argoverse2. The results demonstrate strong robustness and generalization, with notable performance improvements over state-of-the-art approaches and clear ablations confirming the contribution of each component.

Abstract

Currently, high-definition (HD) map construction leans towards a lightweight online generation tendency, which aims to preserve timely and reliable road scene information. However, map elements contain strong shape priors. Subtle and sparse annotations make current detection-based frameworks ambiguous in locating relevant feature scopes and cause the loss of detailed structures in prediction. To alleviate these problems, we propose MGMap, a mask-guided approach that effectively highlights the informative regions and achieves precise map element localization by introducing the learned masks. Specifically, MGMap employs learned masks based on the enhanced multi-scale BEV features from two perspectives. At the instance level, we propose the Mask-activated instance (MAI) decoder, which incorporates global instance and structural information into instance queries by the activation of instance masks. At the point level, a novel position-guided mask patch refinement (PG-MPR) module is designed to refine point locations from a finer-grained perspective, enabling the extraction of point-specific patch information. Compared to the baselines, our proposed MGMap achieves a notable improvement of around 10 mAP for different input modalities. Extensive experiments also demonstrate that our approach showcases strong robustness and generalization capabilities. Our code can be found at https://github.com/xiaolul2/MGMap.

MGMap: Mask-Guided Learning for Online Vectorized HD Map Construction

TL;DR

MGMap tackles the challenge of precise online HD map vectorization under subtle and sparse annotations by introducing mask-guided learning. It combines an Enhanced Multi-Level BEV feature extractor with a Mask-Activated Instance (MAI) decoder and a Position-Guided Mask Patch Refinement (PG-MPR) module to perform coarse instance-level and fine-grained point-level localization. The method jointly leverages learned instance masks and binary mask features to highlight informative regions and refine point locations through ROI-based patch features, achieving substantial mAP gains across nuScenes and Argoverse2. The results demonstrate strong robustness and generalization, with notable performance improvements over state-of-the-art approaches and clear ablations confirming the contribution of each component.

Abstract

Currently, high-definition (HD) map construction leans towards a lightweight online generation tendency, which aims to preserve timely and reliable road scene information. However, map elements contain strong shape priors. Subtle and sparse annotations make current detection-based frameworks ambiguous in locating relevant feature scopes and cause the loss of detailed structures in prediction. To alleviate these problems, we propose MGMap, a mask-guided approach that effectively highlights the informative regions and achieves precise map element localization by introducing the learned masks. Specifically, MGMap employs learned masks based on the enhanced multi-scale BEV features from two perspectives. At the instance level, we propose the Mask-activated instance (MAI) decoder, which incorporates global instance and structural information into instance queries by the activation of instance masks. At the point level, a novel position-guided mask patch refinement (PG-MPR) module is designed to refine point locations from a finer-grained perspective, enabling the extraction of point-specific patch information. Compared to the baselines, our proposed MGMap achieves a notable improvement of around 10 mAP for different input modalities. Extensive experiments also demonstrate that our approach showcases strong robustness and generalization capabilities. Our code can be found at https://github.com/xiaolul2/MGMap.
Paper Structure (23 sections, 14 equations, 11 figures, 9 tables)

This paper contains 23 sections, 14 equations, 11 figures, 9 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. MGMap
  4. BEV Feature Extraction
  5. Mask-Activated Instance Decoder
  6. Position-Guided Mask Patch Refinement
  7. Training Loss
  8. Experiments
  9. Datasets and Benchmarks
  10. Evaluation Metrics
  11. Implementation Details
  12. Main Results
  13. Ablation Study
  14. Conclusion
  15. More Details of Our Method
  16. ...and 8 more sections

Figures (11)

  • Figure 1: For some detailed structures, our proposed MGMap achieves effective map element localization by highlighting the informative regions through the learned masks.
  • Figure 2: Overview of MGMap framework. MGMap mainly consists of three components: (1) BEV Extractor to obtain multi-scale BEV features by transforming from perspective view (PV) to BEV with the enhanced multi-level neck; (2) Mask-Activated Instance (MAI) Decoder is employed to construct and update queries at instance level; (3) Position-Guided Mask Patch Refinement (PG-MPR) module is designed to refine points' positions from local patch features at point level.
  • Figure 3: Illustration of mask constructions at different stages. In MAI decoder, instance masks are generated to activate lane queries, while binary masks are extracted to provide fine-grained patch features in PG-MPR.
  • Figure 4: (a) The conventional deformable attention extracts sparse features from sampling points, which may select irrelevant features; (b) Our proposed Mask Patch Refinement extracts more relevant features from the region of reliable patch.
  • Figure 5: The visual results of MapTR liao2022maptr, our proposed MGMap approach and the corresponding ground truth.
  • ...and 6 more figures