Beyond Ground: Map-Free LiDAR Relocalization for UAVs

TL;DR

This work tackles map-free LiDAR relocalization for UAVs under GNSS-denied conditions, where UAV-specific motions (yaw and altitude changes) degrade vehicle-oriented methods. It introduces MAILS, a pose-regression framework built on Coordinate-Independent Point Cloud Serialization (CIPCS) and Locality-Preserving Sliding Window Attention (LoSWAtt) with yaw- and altitude-invariant positional encoding, enabling robust per-point world-coordinate prediction and RANSAC-based 6-DoF pose estimation. To benchmark realism, the authors construct UAVLoc, a large-scale UAV LiDAR dataset with irregular trajectories and significant altitude variation across four scenes. Empirical results show MAILS achieving state-of-the-art performance on UAVScenes and UAVLoc_U, with ablations confirming the critical roles of constant feature initialization, Softmax-free first LoSWAtt, and position encoding. The work advances practical, map-free UAV relocalization and provides a challenging public dataset to spur further research, potentially enabling reliable on-board localization in challenging environments.

Abstract

Localization is a fundamental capability in unmanned aerial vehicle (UAV) systems. Map-free LiDAR relocalization offers an effective solution for achieving high-precision positioning in environments with weak or unavailable GNSS signals. However, existing LiDAR relocalization methods are primarily tailored to autonomous driving, exhibiting significantly degraded accuracy in UAV scenarios. In this paper, we propose MAILS, a novel map-free LiDAR relocalization framework for UAVs. A Locality-Preserving Sliding Window Attention module is first introduced to extract locally discriminative geometric features from sparse point clouds. To handle substantial yaw rotations and altitude variations encountered during UAV flight, we then design a coordinate-independent feature initialization module and a locally invariant positional encoding mechanism, which together significantly enhance the robustness of feature extraction. Furthermore, existing LiDAR-based relocalization datasets fail to capture real-world UAV flight characteristics, such as irregular trajectories and varying altitudes. To address this gap, we construct a large-scale LiDAR localization dataset for UAVs, which comprises four scenes and various flight trajectories, designed to evaluate UAV relocalization performance under realistic conditions. Extensive experiments demonstrate that our method achieves satisfactory localization precision and consistently outperforms existing techniques by a significant margin. Our code and dataset will be released soon.

Figures (16)

• Figure 1: The challenges of UAV's map-free relocalization compared to vehicle. This paper primarily investigates the map-free UAV relocalization. Compared to map-free vehicle relocalization, UAV relocalization faces greater challenges due to more complex and diverse flight conditions. Thus, developing a robust and accurate algorithm for UAV relocalization is urgently needed.
• Figure 2: Mean position errors of different methods on the UAVScenes and UAVLoc (ours) datasets. Our method achieves the best performance on both datasets.
• Figure 3: The pipeline of MAILS. MAILS consists of two main components:(1) Raw coordinate-independent point cloud serialization initializes point cloud features to a serialization. (2) The Locality-preserving Sliding Window Attention module encodes robust point cloud features (detailed in Fig. \ref{['fig3']}). RANSAC is used to 6-DoF pose estimation in inference stage.
• Figure 4: Locality-Preserving Sliding Window Attention module. This module is designed to encode local geometric features. Notably, the first layer of the LoSWAtt module employs a Softmax-free design to produce differentiated features.
• Figure 5: We scan point clouds and record GT using a UAV equipped with an Ouster OS1-128 LiDAR and a DJI-camera.

Theorems & Definitions (6)

• Definition 1: Local Geometric Token
• Definition 2: Invariant Feature Initialization
• Proposition 1: Yaw--Altitude Invariance of LoSWAtt
• proof
• Theorem 1: Local Yaw--Altitude Invariance of MAILS
• proof