发表机构
Huawei Paris Research Center; Université Paris-Saclay; Univ Evry; L Research（华为巴黎研究中心; 巴黎萨克雷大学; 埃夫里大学; L研究机构）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
HELIOS是一种无需配对图像训练的图像重光照方法，通过反照率条件、反照率蒸馏和GPS太阳角度控制，实现昼夜场景连续重光照，性能优于现有方法。
AI 中文摘要
修改驾驶图像的光照是一项基础挑战，因为大多数数据集是在一天中的特定时间拍摄的。现有方法依赖合成数据或配对的多光照监督，这限制了它们在现实世界多样化且具有挑战性的条件下的泛化能力。为解决该问题，我们提出HELIOS，一种新颖的图像重光照方法，该方法依赖未标记的现实世界数据集，无需任何配对图像进行训练。我们的方法将基于反照率的条件集成到循环一致的扩散流水线中，以防止身份崩溃并确保准确的域迁移。为处理低能见度的夜间条件，我们引入了一种鲁棒的反照率蒸馏策略，该策略从白天域传递结构稳定性。此外，我们用基于GPS衍生太阳角度的细粒度控制机制替换传统文本提示，从而实现昼夜周期内平滑连续的光照操纵。通过广泛的评估和用户研究，我们证明HELIOS在夜间转白天和白天转夜间任务中产生结构一致且逼真的结果，优于最先进的方法。
英文摘要
Modifying the illumination of driving images is a fundamental challenge, as most datasets are captured at specific times of day. Existing methods rely on synthetic data or paired multi-illumination supervision, which limits their generalization to the diverse and challenging conditions of real-world scenarios. To address this, we propose HELIOS, a novel image relighting approach that relies on unlabeled real-world datasets without requiring any paired images for training. Our approach integrates albedo-based conditioning into a cycle-consistent diffusion pipeline to prevent identity collapse and ensure accurate domain translation. To handle low-visibility nighttime conditions, we introduce a robust albedo distillation strategy that transfers structural stability from the daytime domain. Additionally, we replace traditional text prompts with a fine-grained control mechanism based on GPS-derived solar angles, enabling smooth and continuous lighting manipulation across the day-night cycle. Through extensive evaluation and a user study, we demonstrate that HELIOS produces structurally consistent and realistic results in both night-to-day and day-to-night tasks, outperforming state-of-the-art methods.
CommentsProject page: https://hala-djeghim.github.io/HELIOS/
