DiffusionRenderer: Neural Inverse and Forward Rendering with Video Diffusion Models

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Liang, Ruofan, Gojcic, Zan, Ling, Huan, Munkberg, Jacob, Hasselgren, Jon, Lin, Zhi-Hao, Gao, Jun, Keller, Alexander, Vijaykumar, Nandita, Fidler, Sanja, Wang, Zian
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866913753153404928
author Liang, Ruofan
Gojcic, Zan
Ling, Huan
Munkberg, Jacob
Hasselgren, Jon
Lin, Zhi-Hao
Gao, Jun
Keller, Alexander
Vijaykumar, Nandita
Fidler, Sanja
Wang, Zian
author_facet Liang, Ruofan
Gojcic, Zan
Ling, Huan
Munkberg, Jacob
Hasselgren, Jon
Lin, Zhi-Hao
Gao, Jun
Keller, Alexander
Vijaykumar, Nandita
Fidler, Sanja
Wang, Zian
contents Understanding and modeling lighting effects are fundamental tasks in computer vision and graphics. Classic physically-based rendering (PBR) accurately simulates the light transport, but relies on precise scene representations--explicit 3D geometry, high-quality material properties, and lighting conditions--that are often impractical to obtain in real-world scenarios. Therefore, we introduce DiffusionRenderer, a neural approach that addresses the dual problem of inverse and forward rendering within a holistic framework. Leveraging powerful video diffusion model priors, the inverse rendering model accurately estimates G-buffers from real-world videos, providing an interface for image editing tasks, and training data for the rendering model. Conversely, our rendering model generates photorealistic images from G-buffers without explicit light transport simulation. Experiments demonstrate that DiffusionRenderer effectively approximates inverse and forwards rendering, consistently outperforming the state-of-the-art. Our model enables practical applications from a single video input--including relighting, material editing, and realistic object insertion.
format Preprint
id arxiv_https___arxiv_org_abs_2501_18590
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle DiffusionRenderer: Neural Inverse and Forward Rendering with Video Diffusion Models
Liang, Ruofan
Gojcic, Zan
Ling, Huan
Munkberg, Jacob
Hasselgren, Jon
Lin, Zhi-Hao
Gao, Jun
Keller, Alexander
Vijaykumar, Nandita
Fidler, Sanja
Wang, Zian
Computer Vision and Pattern Recognition
Graphics
Understanding and modeling lighting effects are fundamental tasks in computer vision and graphics. Classic physically-based rendering (PBR) accurately simulates the light transport, but relies on precise scene representations--explicit 3D geometry, high-quality material properties, and lighting conditions--that are often impractical to obtain in real-world scenarios. Therefore, we introduce DiffusionRenderer, a neural approach that addresses the dual problem of inverse and forward rendering within a holistic framework. Leveraging powerful video diffusion model priors, the inverse rendering model accurately estimates G-buffers from real-world videos, providing an interface for image editing tasks, and training data for the rendering model. Conversely, our rendering model generates photorealistic images from G-buffers without explicit light transport simulation. Experiments demonstrate that DiffusionRenderer effectively approximates inverse and forwards rendering, consistently outperforming the state-of-the-art. Our model enables practical applications from a single video input--including relighting, material editing, and realistic object insertion.
title DiffusionRenderer: Neural Inverse and Forward Rendering with Video Diffusion Models
topic Computer Vision and Pattern Recognition
Graphics
url https://arxiv.org/abs/2501.18590