UrbanIR: Large-Scale Urban Scene Inverse Rendering from a Single Video

Zhi-Hao Lin, Bohan Liu, Yi-Ting Chen, David Forsyth, Jia-Bin Huang, Anand Bhattad, Shenlong Wang

Keywords: Inverse rendering, neural radiance field, outdoor scene

TL;DR: UrbanIR decompose scene properties and perform view synthesis, relighting, and object insertion from single video under same illumination

Abstract: We present UrbanIR~(Urban Scene Inverse Rendering), a new inverse graphics model that enables realistic, free-viewpoint renderings of scenes under various lighting conditions with a single video. It accurately infers shape, albedo, visibility, and sun and sky illumination from wide-baseline videos, such as those from car-mounted cameras, differing from NeRF's dense view settings. In this context, standard methods often yield subpar geometry and material estimates, such as inaccurate roof representations and numerous `floaters'. UrbanIR addresses these issues with novel losses that reduce errors in inverse graphics inference and rendering artifacts. Its techniques allow for precise shadow volume estimation in the original scene. The model's outputs support controllable editing, enabling photorealistic free-viewpoint renderings of night simulations, relit scenes, and inserted objects, marking a significant improvement over existing state-of-the-art methods. Our code and data will be made publicly available upon acceptance.

Submission Number: 33