提升自动驾驶天气模拟的传感器一致性，让仿真更贴近真实。
Multi-Sensor Alignment for Weather Simulations
- 提出ReDAM和Unified-weather-edit方法，分别对齐雾、雨雪的强度与粒子分布。
- 非对齐仿真使3D检测模型性能虚高，对齐后结果更真实可靠。
- 可直接用于现有融合模型微调，提升恶劣天气下目标检测鲁棒性。
自动驾驶感知任务需在恶劣天气下正常工作，但真实世界天气数据集稀缺，因此天气模拟成为重要替代方案。为确保模拟数据与真实气象特征一致，包括天气严重程度和多传感器间粒子位置分布，本文提出参考数据集对齐方法（ReDAM）实现雾天强度对齐，以及受Weather-edit启发的Unified-weather-edit方法实现雨雪天气粒子位置对齐。通过统计与几何测试验证了两种方法的有效性。实验发现，未对齐版本的3D检测模型表现明显过于乐观，而对齐后的多传感器仿真能有效提升现有传感器融合模型在3D目标检测任务中的鲁棒性。
原文摘要 · Abstract (English)
Perception tasks for autonomous vehicles need to work satisfactorily in adverse weather conditions. Due to lack of real-world weather datasets, weather simulations are a promising alternative. To ensure simulations closely mirror real-world weather data, it's crucial that they represent the same weather characteristics, including severity and particle positioning, across different sensors. To achieve this, we propose the Reference Dataset Alignment Method (ReDAM) for weather intensity alignment in fog and Unified-weather-edit (inspired by Weather-edit[1]) for particle positioning alignment in rain and snow. We validate both alignment methods using statistical and geometrical tests, respectively. We find that 3D detection models for non-aligned versions tend to be overly optimistic as compared to aligned versions. We also show the aligned-multi-sensor simulation's effectiveness for achieving robustness for 3D object detection task by finetuning existing sensor fusion models on it.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。