Generalizing Unsupervised Lidar Odometry Model from Normal to Snowy Weather Conditions

2025-09-02

AI Summary

Unsupervised LiDAR odometry suffers from poor generalization and unstable pose estimation under adverse weather (e.g., snowfall) due to snow-induced noise. To address this, we propose a robust denoising and local feature enhancement framework. Our method introduces: (1) a Patch Spatial Measure module that automatically detects discrete snow noise; (2) an adaptive Patch Point Weight Predictor to enhance local point cloud discriminability; and (3) a fusion of intensity-threshold masking and multimodal features for efficient snow noise suppression in fully unsupervised settings. Crucially, the model is trained exclusively on clear-weather data yet achieves significantly improved pose estimation accuracy and stability under snowy and dynamic conditions. It demonstrates strong cross-weather generalization while maintaining real-time inference capability. This work provides a novel and practical solution for reliable autonomous vehicle localization across varying climatic conditions.

Technology Category

• Intelligent Robots: State Estimation
• Machine Learning: Unsupervised & Self-Supervised Learning
• Computer Vision: Adversarial Attacks & Robustness

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Deep learning-based LiDAR odometry is crucial for autonomous driving and robotic navigation, yet its performance under adverse weather, especially snowfall, remains challenging. Existing models struggle to generalize across conditions due to sensitivity to snow-induced noise, limiting real-world use. In this work, we present an unsupervised LiDAR odometry model to close the gap between clear and snowy weather conditions. Our approach focuses on effective denoising to mitigate the impact of snowflake noise and outlier points on pose estimation, while also maintaining computational efficiency for real-time applications. To achieve this, we introduce a Patch Spatial Measure (PSM) module that evaluates the dispersion of points within each patch, enabling effective detection of sparse and discrete noise. We further propose a Patch Point Weight Predictor (PPWP) to assign adaptive point-wise weights, enhancing their discriminative capacity within local regions. To support real-time performance, we first apply an intensity threshold mask to quickly suppress dense snowflake clusters near the LiDAR, and then perform multi-modal feature fusion to refine the point-wise weight prediction, improving overall robustness under adverse weather. Our model is trained in clear weather conditions and rigorously tested across various scenarios, including snowy and dynamic. Extensive experimental results confirm the effectiveness of our method, demonstrating robust performance in both clear and snowy weather. This advancement enhances the model's generalizability and paves the way for more reliable autonomous systems capable of operating across a wider range of environmental conditions.

Problem

• Generalizing LiDAR odometry from normal to snowy weather conditions
• Mitigating snow-induced noise and outlier points in LiDAR data
• Maintaining real-time computational efficiency for autonomous navigation

Innovation

• Patch Spatial Measure for noise detection
• Patch Point Weight Predictor for adaptive weighting
• Intensity threshold mask with multi-modal fusion