GAINS: Gaussian-based Inverse Rendering from Sparse Multi-View Captures

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Noras, Patrick, Choi, Jun Myeong, Stricker, Didier, Peers, Pieter, Sengupta, Roni
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866917137483825152
author Noras, Patrick
Choi, Jun Myeong
Stricker, Didier
Peers, Pieter
Sengupta, Roni
author_facet Noras, Patrick
Choi, Jun Myeong
Stricker, Didier
Peers, Pieter
Sengupta, Roni
contents Recent advances in Gaussian Splatting-based inverse rendering extend Gaussian primitives with shading parameters and physically grounded light transport, enabling high-quality material recovery from dense multi-view captures. However, these methods degrade sharply under sparse-view settings, where limited observations lead to severe ambiguity between geometry, reflectance, and lighting. We introduce GAINS (Gaussian-based Inverse rendering from Sparse multi-view captures), a two-stage inverse rendering framework that leverages learning-based priors to stabilize geometry and material estimation. GAINS first refines geometry using monocular depth/normal and diffusion priors, then employs segmentation, intrinsic image decomposition (IID), and diffusion priors to regularize material recovery. Extensive experiments on synthetic and real-world datasets show that GAINS significantly improves material parameter accuracy, relighting quality, and novel-view synthesis compared to state-of-the-art Gaussian-based inverse rendering methods, especially under sparse-view settings. Project page: https://patrickbail.github.io/gains/
format Preprint
id arxiv_https___arxiv_org_abs_2512_09925
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle GAINS: Gaussian-based Inverse Rendering from Sparse Multi-View Captures
Noras, Patrick
Choi, Jun Myeong
Stricker, Didier
Peers, Pieter
Sengupta, Roni
Computer Vision and Pattern Recognition
Recent advances in Gaussian Splatting-based inverse rendering extend Gaussian primitives with shading parameters and physically grounded light transport, enabling high-quality material recovery from dense multi-view captures. However, these methods degrade sharply under sparse-view settings, where limited observations lead to severe ambiguity between geometry, reflectance, and lighting. We introduce GAINS (Gaussian-based Inverse rendering from Sparse multi-view captures), a two-stage inverse rendering framework that leverages learning-based priors to stabilize geometry and material estimation. GAINS first refines geometry using monocular depth/normal and diffusion priors, then employs segmentation, intrinsic image decomposition (IID), and diffusion priors to regularize material recovery. Extensive experiments on synthetic and real-world datasets show that GAINS significantly improves material parameter accuracy, relighting quality, and novel-view synthesis compared to state-of-the-art Gaussian-based inverse rendering methods, especially under sparse-view settings. Project page: https://patrickbail.github.io/gains/
title GAINS: Gaussian-based Inverse Rendering from Sparse Multi-View Captures
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2512.09925