Weather-Conditioned Branch Routing for Robust LiDAR-Radar 3D Object Detection

2026-04-06

AI Summary

This work addresses the challenge of degraded sensor reliability under adverse weather conditions and the limited adaptability of existing LiDAR–4D radar fusion methods. To overcome this, the authors propose a weather-condition-driven dynamic branch routing mechanism that maintains three parallel feature streams—LiDAR-only, 4D radar-only, and conditionally gated fusion—and employs a lightweight router to softly aggregate them using visual and semantic cues. Weather-supervised learning combined with diversity regularization is introduced to prevent branch collapse and enable interpretable, context-aware modality preference switching. Evaluated on the K-Radar benchmark, the method significantly enhances the robustness of 3D object detection in harsh weather, achieving state-of-the-art performance while exhibiting clear and explainable modality selection behavior.

Abstract

Robust 3D object detection in adverse weather is highly challenging due to the varying reliability of different sensors. While existing LiDAR-4D radar fusion methods improve robustness, they predominantly rely on fixed or weakly adaptive pipelines, failing to dy-namically adjust modality preferences as environmental conditions change. To bridge this gap, we reformulate multi-modal perception as a weather-conditioned branch routing problem. Instead of computing a single fused output, our framework explicitly maintains three parallel 3D feature streams: a pure LiDAR branch, a pure 4D radar branch, and a condition-gated fusion branch. Guided by a condition token extracted from visual and semantic prompts, a lightweight router dynamically predicts sample-specific weights to softly aggregate these representations. Furthermore, to prevent branch collapse, we introduce a weather-supervised learning strategy with auxiliary classification and diversity regularization to enforce distinct, condition-dependent routing behaviors. Extensive experiments on the K-Radar benchmark demonstrate that our method achieves state-of-the-art performance. Furthermore, it provides explicit and highly interpretable insights into modality preferences, transparently revealing how adaptive routing robustly shifts reliance between LiDAR and 4D radar across diverse adverse-weather scenarios. The source code with be released.

Problem

• adverse weather
• LiDAR-radar fusion
• 3D object detection
• modality reliability
• dynamic routing

Innovation

• weather-conditioned routing
• LiDAR-radar fusion
• adaptive multimodal perception
• branch diversity regularization
• 3D object detection