🤖 AI Summary
To address the fundamental trade-off between accuracy and robustness in LiDAR SLAM under dynamic object interference, point cloud noise, and unstructured environments, this paper proposes a neural-descriptor-based adaptive noise filtering SLAM framework. Our method jointly identifies dynamic points and low-quality points via a dedicated dynamic segmentation head and a global importance scoring head. To enhance geometric consistency modeling, we introduce a cross-layer graph convolutional module (GLI-GCN) that fuses multi-scale neighborhood structures. Furthermore, the framework enables end-to-end adaptive selection of high-contribution feature points and simultaneous noise suppression. Extensive experiments on multiple public benchmarks demonstrate significant improvements: average absolute trajectory error (ATE) is reduced by 23.6%, and system robustness is substantially enhanced—particularly in highly dynamic and strongly noisy scenarios. The proposed approach thus achieves superior localization accuracy without compromising reliability in challenging real-world conditions.
📝 Abstract
LiDAR SLAM has demonstrated significant application value in various fields, including mobile robot navigation and high-precision map construction. However, existing methods often need to make a trade-off between positioning accuracy and system robustness when faced with dynamic object interference, point cloud noise, and unstructured environments. To address this challenge, we propose an adaptive noise filtering SLAM strategy-ADA-DPM, achieving excellent preference in both aspects. We design the Dynamic Segmentation Head to predict the category of feature points belonging to dynamic points, to eliminate dynamic feature points; design the Global Importance Scoring Head to adaptively select feature points with higher contribution and features while suppressing noise interference; and construct the Cross Layer Intra-Graph Convolution Module (GLI-GCN) to fuse multi-scale neighborhood structures, thereby enhancing the discriminative ability of overlapping features. Finally, to further validate the effectiveness of our method, we tested it on several publicly available datasets and achieved outstanding results.