发表机构
Huawei Paris Research Center; Gustave Eiffel University; IGN-ENSG（华为巴黎研究中心; 古斯塔夫·埃菲尔大学; 法国国家地理信息与森林和环境信息研究所）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
针对自动驾驶系统在不同天气条件下可靠感知的挑战，提出Cyclone框架，基于潜在扩散，利用循环一致约束和图像-文本模型知识，无需配对数据生成多种天气条件，实验表明其输出更优，还可提炼为视频扩散模型。
AI 中文摘要
在不同天气条件下的可靠感知仍然是自动驾驶系统的一个主要挑战。一种提高鲁棒性的常见策略是为训练感知模型合成不利天气条件，或应用天气去除技术来恢复干净的输入。然而，现有方法通常依赖于合成数据增强或基于物理的特定任务模型，这些模型需要配对的训练数据，并且往往难以生成逼真的天气效果或稳健地推广到域外场景。针对这个问题，我们提出了Cyclone，一个基于潜在扩散的天气编辑统一框架，配备了循环一致约束和来自图像-文本模型的知识。Cyclone能够在不同场景中生成多种天气条件，同时无需配对数据。实验结果表明，我们的方法比现有基线产生更逼真、保留结构的输出，并在几个下游驾驶感知任务中带来一致的改进。此外，我们证明Cyclone可以提炼为一个用于时间一致天气编辑的视频扩散模型。
英文摘要
Reliable perception under diverse weather conditions remains a major challenge for autonomous driving systems. A common strategy to improve robustness is either to synthesize adverse weather conditions for training perception models or to apply weather-removal techniques to recover clean inputs. However, existing approaches typically rely on synthetic data augmentation or physics-based, task-specific models that require paired training data and often struggle to generate realistic weather effects or generalize robustly to out-of-domain scenarios. Toward this problem, we present Cyclone, a unified framework for weather editing based on latent diffusion, equipped with cycle-consistent constraints and knowledge from image-text models. Cyclone enables the generation of multiple weather conditions across diverse scenes while eliminating the need for paired data. Experimental results show that our approach produces more realistic, structure-preserving outputs than existing baselines and leads to consistent improvements across several downstream driving perception tasks. Furthermore, we demonstrate that Cyclone can be distilled to a video diffusion model for temporally consistent weather editing.
CommentsProject page: https://ntaquan0125.github.io/weather-cyclone/
