Physically Controllable Relighting of Photographs
This work addresses the challenge of applying explicit physical lighting control, common in 3D graphics tools, to real-world images for users in computer vision and graphics.
The paper tackles the problem of enabling physically controllable relighting of in-the-wild photographs by combining traditional rendering with neural rendering, achieving photorealistic results through a self-supervised approach trained on raw image collections.
We present a self-supervised approach to in-the-wild image relighting that enables fully controllable, physically based illumination editing. We achieve this by combining the physical accuracy of traditional rendering with the photorealistic appearance made possible by neural rendering. Our pipeline works by inferring a colored mesh representation of a given scene using monocular estimates of geometry and intrinsic components. This representation allows users to define their desired illumination configuration in 3D. The scene under the new lighting can then be rendered using a path-tracing engine. We send this approximate rendering of the scene through a feed-forward neural renderer to predict the final photorealistic relighting result. We develop a differentiable rendering process to reconstruct in-the-wild scene illumination, enabling self-supervised training of our neural renderer on raw image collections. Our method represents a significant step in bringing the explicit physical control over lights available in typical 3D computer graphics tools, such as Blender, to in-the-wild relighting.