SparseGS-W: Sparse-View 3D Gaussian Splatting in the Wild with Generative Priors

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Li, Yiqing, Wang, Xuan, Wu, Jiawei, Ma, Yikun, Jin, Zhi
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914292330135552
author Li, Yiqing
Wang, Xuan
Wu, Jiawei
Ma, Yikun
Jin, Zhi
author_facet Li, Yiqing
Wang, Xuan
Wu, Jiawei
Ma, Yikun
Jin, Zhi
contents Synthesizing novel views of large-scale scenes from unconstrained in-the-wild images is an important but challenging task in computer vision. Existing methods, which optimize per-image appearance and transient occlusion through implicit neural networks from dense training views (approximately 1000 images), struggle to perform effectively under sparse input conditions, resulting in noticeable artifacts. To this end, we propose SparseGS-W, a novel framework based on 3D Gaussian Splatting that enables the reconstruction of complex outdoor scenes and handles occlusions and appearance changes with as few as five training images. We leverage geometric priors and constrained diffusion priors to compensate for the lack of multi-view information from extremely sparse input. Specifically, we propose a plug-and-play Constrained Novel-View Enhancement module to iteratively improve the quality of rendered novel views during the Gaussian optimization process. Furthermore, we propose an Occlusion Handling module, which flexibly removes occlusions utilizing the inherent high-quality inpainting capability of constrained diffusion priors. Both modules are capable of extracting appearance features from any user-provided reference image, enabling flexible modeling of illumination-consistent scenes. Extensive experiments on the PhotoTourism and Tanks and Temples datasets demonstrate that SparseGS-W achieves state-of-the-art performance not only in full-reference metrics, but also in commonly used non-reference metrics such as FID, ClipIQA, and MUSIQ.
format Preprint
id arxiv_https___arxiv_org_abs_2503_19452
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle SparseGS-W: Sparse-View 3D Gaussian Splatting in the Wild with Generative Priors
Li, Yiqing
Wang, Xuan
Wu, Jiawei
Ma, Yikun
Jin, Zhi
Computer Vision and Pattern Recognition
Synthesizing novel views of large-scale scenes from unconstrained in-the-wild images is an important but challenging task in computer vision. Existing methods, which optimize per-image appearance and transient occlusion through implicit neural networks from dense training views (approximately 1000 images), struggle to perform effectively under sparse input conditions, resulting in noticeable artifacts. To this end, we propose SparseGS-W, a novel framework based on 3D Gaussian Splatting that enables the reconstruction of complex outdoor scenes and handles occlusions and appearance changes with as few as five training images. We leverage geometric priors and constrained diffusion priors to compensate for the lack of multi-view information from extremely sparse input. Specifically, we propose a plug-and-play Constrained Novel-View Enhancement module to iteratively improve the quality of rendered novel views during the Gaussian optimization process. Furthermore, we propose an Occlusion Handling module, which flexibly removes occlusions utilizing the inherent high-quality inpainting capability of constrained diffusion priors. Both modules are capable of extracting appearance features from any user-provided reference image, enabling flexible modeling of illumination-consistent scenes. Extensive experiments on the PhotoTourism and Tanks and Temples datasets demonstrate that SparseGS-W achieves state-of-the-art performance not only in full-reference metrics, but also in commonly used non-reference metrics such as FID, ClipIQA, and MUSIQ.
title SparseGS-W: Sparse-View 3D Gaussian Splatting in the Wild with Generative Priors
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2503.19452