Rethinking Range-View LiDAR Segmentation in Adverse Weather

2025-06-10

AI Summary

To address the poor generalization of range-view LiDAR semantic segmentation under adverse weather conditions (e.g., rain, fog, snow), this paper proposes a lightweight, modular robust segmentation framework that requires no modification to the backbone network. Methodologically, it introduces: (1) a novel dual-branch stem architecture—Geometric Anomaly Suppression (GAS) and Reflectance Distortion Calibration (RDC); (2) Memory-Guided Adaptive Instance Normalization (MGAIN) for dynamic reflectance correction; and (3) a synergistic design integrating dual-stream feature disentanglement, geometric noise suppression, and multi-scale feature fusion. Evaluated on multiple adverse-weather benchmarks, the framework achieves an average mIoU improvement of 4.2% with less than 3% additional inference overhead, enabling real-time deployment.

Technology Category

• Computer Vision: Segmentation
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Natural Language Processing: Safety and 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: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

LiDAR segmentation has emerged as an important task to enrich multimedia experiences and analysis. Range-view-based methods have gained popularity due to their high computational efficiency and compatibility with real-time deployment. However, their generalized performance under adverse weather conditions remains underexplored, limiting their reliability in real-world environments. In this work, we identify and analyze the unique challenges that affect the generalization of range-view LiDAR segmentation in severe weather. To address these challenges, we propose a modular and lightweight framework that enhances robustness without altering the core architecture of existing models. Our method reformulates the initial stem block of standard range-view networks into two branches to process geometric attributes and reflectance intensity separately. Specifically, a Geometric Abnormality Suppression (GAS) module reduces the influence of weather-induced spatial noise, and a Reflectance Distortion Calibration (RDC) module corrects reflectance distortions through memory-guided adaptive instance normalization. The processed features are then fused and passed to the original segmentation pipeline. Extensive experiments on different benchmarks and baseline models demonstrate that our approach significantly improves generalization to adverse weather with minimal inference overhead, offering a practical and effective solution for real-world LiDAR segmentation.

Problem

• Improving LiDAR segmentation robustness in adverse weather
• Addressing weather-induced noise and reflectance distortions
• Enhancing range-view methods without altering core architectures

Innovation

• Modular lightweight framework enhances robustness
• Dual-branch stem processes geometry and reflectance
• GAS and RDC modules reduce weather noise