# Open-Weather Robust 3D Detection via Dual-Critic Diffusion Alignment Source: [Open-Weather Robust 3D Detection via Dual-Critic Diffusion Alignment]() Collection index: [llms.txt]() Data updated: 2026-07-08T05:00:13.057Z Authors: Shuyao Li, Chuanxing Geng, Heyang Sun, Qiang Zhou, Jingjing Gu Published: 2026-07-02T10:13:41.000Z ## Abstract Robust 3D object detection under adverse weather remains a critical hurdle for autonomous driving. Despite progress with LiDAR-4D radar fusion, most methods are constrained by a closed-world assumption, implicitly requiring training and test weather to align in both type and severity. This premise fails in practice: the open-ended nature of weather, and even variations within a single type like rain, cause dramatically different LiDAR degradation patterns, leading to significant performance drops in unseen conditions. To address this, we present Dual-Critic Guided Diffusion Alignment (DCDA), a weather-agnostic framework that learns to recover degraded LiDAR features toward a clean manifold. Rather than modeling specific weather types, DCDA employs a 4D radar-conditioned diffusion process to progressively refine features, guided by two complementary critics. (i) A detection-guided critic, anchored by a pre-trained clean-weather model, ensures that the refined features retain object-level discriminability and localization accuracy. (ii) A weather adversarial critic enforces holistic distributional consistency with clean-weather representations. By aligning features through semantic and distributional constraints rather than explicit weather modeling, DCDA generalizes effectively to unseen weather types and severities without requiring paired data or weather labels. We further introduce a structured open-weather benchmark with held-out type-severity combinations and extensive experiments verify DCDA's advantages. ## Recorded benchmark results A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. - [3D Object Detection on K-Radar (Normal)](): 3D Object Detection; K-Radar (Normal); BEV AP: 73.3. - [3D Object Detection on K-Radar (non-Normal)](): 3D Object Detection; K-Radar (non-Normal); BEV AP: 58.8. - [3D Object Detection on K-Radar (type+severity-open)](): 3D Object Detection; K-Radar (type+severity-open); BEV AP (Fog, Low Severity): 68.68. - [3D Object Detection on K-Radar synthetic severity-open, Fog, Light](): 3D Object Detection; K-Radar synthetic severity-open, Fog, Light; BEV AP: 68.81. - [3D Object Detection on K-Radar synthetic severity-open, Fog, Heavy](): 3D Object Detection; K-Radar synthetic severity-open, Fog, Heavy; BEV AP: 35.16. - [3D Object Detection on K-Radar synthetic severity-open, Snow, Light](): 3D Object Detection; K-Radar synthetic severity-open, Snow, Light; BEV AP: 42.35. - [3D Object Detection on K-Radar synthetic severity-open Snow Medium](): 3D Object Detection; K-Radar synthetic severity-open Snow Medium; BEV AP: 44.27. - [3D Object Detection on K-Radar synthetic severity-open Wet Ground Medium](): 3D Object Detection; K-Radar synthetic severity-open Wet Ground Medium; BEV AP: 67.4. - [3D Object Detection on K-Radar synthetic severity-open, Wet Ground, Heavy](): 3D Object Detection; K-Radar synthetic severity-open, Wet Ground, Heavy; BEV AP: 67.37. - [3D Object Detection on K-Radar synthetic severity-open, Wet Ground, Light](): 3D Object Detection; K-Radar synthetic severity-open, Wet Ground, Light; BEV AP: 75.13. - [3D Object Detection on K-Radar synthetic severity-open, Snow, Heavy](): 3D Object Detection; K-Radar synthetic severity-open, Snow, Heavy; BEV AP: 32.77. - [3D Object Detection on K-Radar (test)](): 3D Object Detection; K-Radar (test); AP BEV (Normal): 68.39. ## Original sources and implementation [Original paper]() [PDF]()