发表机构
Université Laval; Adobe Research; Computer Vision Center; Universitat Autònoma de Barcelona（拉瓦尔大学; Adobe研究院; 计算机视觉中心; 巴塞罗那自治大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本工作提出一个基准，通过测试生成模型在真实照片中插入物体时保持光照一致性的能力，来评估其光照理解，并定量估计光照方向、颜色和辐射分布以衡量准确性。
AI 中文摘要
精确的照明建模对于逼真的图像合成和场景理解至关重要。然而，目前很少有研究探讨图像生成模型是否擅长这一任务，或者物理合理性是否仍然是它们面临的关键挑战。显然，在逼真图像合成方面已经取得了显著进展，但模型是否真正以物理准确的方式理解光照呢？为了回答这个问题，本工作提出了一个基准，用于评估生成模型的光照理解和协调能力。我们的关键见解是，评估此类模型的光照理解只需测试它们如何将新物体插入真实照片中，同时保持一致的照明。为此，我们使用一个多光照数据集，其中包含作为“光探针”的简单物体的图像，并提示模型将同一物体修复到原始图像上，然后将生成结果与真实光探针进行比较。然后，我们从修复的探针中估计光照方向、颜色和辐射分布，从而提供光照准确性和光度真实性的定量度量。我们的工作建立了一个可扩展的评估协议，以系统评估生成模型如何捕捉和再现真实世界的光照，为基准测试任何未来模型的光度准确性提供了基础。所有代码和数据可在以下网址获取：此 https URL。
英文摘要
Accurate modelling of illumination is central to realistic image synthesis and scene understanding. Yet, there is little exploration into whether image generative models are good at this task or whether physical plausibility remains a key challenge for them. Clearly, significant progress has been made in realistic image synthesis, but do models truly understand lighting in a physically accurate manner? To answer this question, this work proposes a benchmark to assess the lighting understanding and harmonisation capabilities of generative models. Our key insight is that evaluating lighting understanding for such models only requires testing how well they insert novel objects into real photographs whilst maintaining consistent illumination. To do so, we use a multi-illumination dataset with images containing simple objects serving as ``light probes'', and prompt models to inpaint the same object onto the original image, then compare the generated results against the ground-truth light probes. We then estimate the lighting direction, colour and radiance distribution from the inpainted probes, providing a quantitative measure of illumination accuracy and photometric realism. Our work establishes a scalable evaluation protocol to systematically assess how well generative models capture and reproduce real-world lighting, offering a foundation for benchmarking the photometric accuracy of any future models. All code and data are available at https://lvsn.github.io/SheddingLight/ .
CommentsAccepted to ACM Transactions on Graphics (SIGGRAPH Asia 2026), vol. 45, no. 6, article 227, December 2026. 25 pages. Project page: https://lvsn.github.io/SheddingLight/
DOI：10.1145/3842579
