提出3D道路拓扑理解新模型，提升精度与泛化能力
TopoMaskV3: 3D Mask Head with Dense Offset and Height Predictions for Road Topology Understanding
- 引入密集偏移场和高度图，实现3D道路中心线直接预测
- 在地理隔离数据集上达到28.5的OLS最优性能
- 解决地理数据泄露问题，适合高精地图与自动驾驶场景
基于掩码的道路拓扑理解方法（如TopoMaskV2）通过密集栅格中间表示生成中心线，是查询式方法的补充。但此前工作仅限于2D预测，存在严重离散化伪影，需融合参数化头。本文提出TopoMaskV3，通过两个新型密集预测头——密集偏移场用于在现有鸟瞰图分辨率下进行亚像素级修正，密集高度图用于直接3D估计，将该流程推进为稳健的独立3D预测器。此外，首次在道路拓扑评估中解决地理数据泄露问题，提出(1)地理上互不重叠的数据划分以防止记忆化并确保公平泛化，(2)长距离（±100米）基准测试。TopoMaskV3在该地理分离基准上取得28.5的最优OLS分数，超越所有先前方法。分析表明，掩码表示对地理过拟合更具鲁棒性，而激光雷达融合在远距离表现最佳，且在重叠原始划分上相对增益更大，提示重叠区域存在记忆化效应。
原文摘要 · Abstract (English)
Mask-based paradigms for road topology understanding, such as TopoMaskV2, offer a complementary alternative to query-based methods by generating centerlines via a dense rasterized intermediate representation. However, prior work was limited to 2D predictions and suffered from severe discretization artifacts, necessitating fusion with parametric heads. We introduce TopoMaskV3, which advances this pipeline into a robust, standalone 3D predictor via two novel dense prediction heads: a dense offset field for sub-grid discretization correction within the existing BEV resolution, and a dense height map for direct 3D estimation. Beyond the architecture, we are the first to address geographic data leakage in road topology evaluation by introducing (1) geographically distinct splits to prevent memorization and ensure fair generalization, and (2) a long-range (+/-100 m) benchmark. TopoMaskV3 achieves state-of-the-art 28.5 OLS on this geographically disjoint benchmark, surpassing all prior methods. Our analysis shows that the mask representation is more robust to geographic overfitting than Bezier, while LiDAR fusion is most beneficial at long range and exhibits larger relative gains on the overlapping original split, suggesting overlap-induced memorization effects.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。