Hierarchical Pose Estimation and Mapping with Multi-Scale Neural Feature Fields

2024-12-11

International Conference on Robotic Computing

AI Summary

Neural implicit SLAM for large-scale outdoor scenes faces two key challenges: unknown sensor poses and difficulty modeling long-sequence LiDAR data. To address these, this paper proposes a hierarchical pose estimation framework coupled with multi-scale neural feature field modeling. We innovatively formulate implicit mapping from a probabilistic perspective, enabling structured sparse representation suitable for large-scale environments. Additionally, we design a hierarchical pose optimization network to enhance localization robustness and mapping stability over extended trajectories. Extensive evaluations on the KITTI and MaiCity datasets demonstrate that our method achieves state-of-the-art pose estimation accuracy under fully unknown initial poses, while significantly outperforming existing baselines in map completeness and long-term stability.

Technology Category

• Intelligent Robots: State Estimation
• Search and Optimization: Learning to Search
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Robotic applications require a comprehensive understanding of the scene. In recent years, neural fields-based approaches that parameterize the entire environment have become popular. These approaches are promising due to their continuous nature and their ability to learn scene priors. However, the use of neural fields in robotics becomes challenging when dealing with unknown sensor poses and sequential measurements. This paper focuses on the problem of sensor pose estimation for large-scale neural implicit SLAM. We investigate implicit mapping from a probabilistic perspective and propose hierarchical pose estimation with a corresponding neural network architecture. Our method is well-suited for large-scale implicit map representations. The proposed approach operates on consecutive outdoor LiDAR scans and achieves accurate pose estimation, while maintaining stable mapping quality for both short and long trajectories. We built our method on a structured and sparse implicit representation suitable for large-scale reconstruction and evaluated it using the KITTI and MaiCity datasets. Our approach outperforms the baseline in terms of mapping with unknown poses and achieves state-of-the-art localization accuracy.

Problem

• Neural Fields
• Robot Localization
• Sensor Uncertainty

Innovation

• Multi-level Features
• Neural Networks
• High-precision Localization