DeepMapping2: Self-Supervised Large-Scale LiDAR Map Optimization
DeepMapping2: Self-Supervised Large-Scale LiDAR Map Optimization
复制标题
DOI：
10.1109/cvpr52729.2023.00898
复制
发表时间：
2022-12
期刊：
影响因子：
--
通讯作者：
Chao Chen;Xinhao Liu;Yiming Li;Li Ding;Chen Feng-
中科院分区：
文献类型：
--
作者：
Chao Chen;Xinhao Liu;Yiming Li;Li Ding;Chen Feng-
激光雷达测绘对于自动驾驶和移动机器人技术来说很重要，但也具有挑战性。为了解决这样的全局点云配准问题，DeepMapping [1] 将复杂的地图估计转换为简单深度网络的自监督训练。尽管 DeepMapping 在小型数据集上具有广泛的收敛范围，但在具有数千帧的大规模数据集上仍然无法产生令人满意的结果。这是由于缺乏闭环和精确的跨帧点对应，以及其全球定位网络收敛缓慢。我们提出 DeepMapping2，通过添加两种新技术来解决这些问题：（1）基于循环闭合的地图拓扑组织训练批次，以及（2）利用成对配准的自监督局部到全局点一致性损失。我们在 KITTI、NCLT 和 Nebula 等公共数据集上的实验和消融研究证明了我们方法的有效性。
LiDAR mapping is important yet challenging in selfdriving and mobile robotics. To tackle such a global point cloud registration problem, DeepMapping [1] converts the complex map estimation into a self-supervised training of simple deep networks. Despite its broad convergence range on small datasets, DeepMapping still cannot produce satisfactory results on large-scale datasets with thousands of frames. This is due to the lack of loop closures and exact cross-frame point correspondences, and the slow convergence of its global localization network. We propose DeepMapping2 by adding two novel techniques to address these issues: (1) organization of training batch based on map topology from loop closing, and (2) self-supervised local-to-global point consistency loss leveraging pairwise registration. Our experiments and ablation studies on public datasets such as KITTI, NCLT, and Nebula demonstrate the effectiveness of our method.