GSFix3D: Diffusion-Guided Repair of Novel Views in Gaussian Splatting

TL;DR

GSFix3D introduces a diffusion-guided framework that enhances 3D Gaussian Splatting for high-quality novel view synthesis, especially in challenging scenarios with limited data or extreme viewpoints.

Contribution

The paper presents GSFix3D and GSFixer, novel methods that incorporate diffusion models into 3D reconstruction, enabling robust view repair and inpainting with minimal fine-tuning.

Findings

Achieves state-of-the-art results on challenging benchmarks.

Demonstrates robustness to pose errors in real-world tests.

Requires minimal scene-specific fine-tuning.

Abstract

Recent developments in 3D Gaussian Splatting have significantly enhanced novel view synthesis, yet generating high-quality renderings from extreme novel viewpoints or partially observed regions remains challenging. Meanwhile, diffusion models exhibit strong generative capabilities, but their reliance on text prompts and lack of awareness of specific scene information hinder accurate 3D reconstruction tasks. To address these limitations, we introduce GSFix3D, a novel framework that improves the visual fidelity in under-constrained regions by distilling prior knowledge from diffusion models into 3D representations, while preserving consistency with observed scene details. At its core is GSFixer, a latent diffusion model obtained via our customized fine-tuning protocol that can leverage both mesh and 3D Gaussians to adapt pretrained generative models to a variety of environments and artifact types from different reconstruction methods, enabling robust novel view repair for unseen camera poses. Moreover, we propose a random mask augmentation strategy that empowers GSFixer to plausibly inpaint missing regions. Experiments on challenging benchmarks demonstrate that our GSFix3D and GSFixer achieve state-of-the-art performance, requiring only minimal scene-specific fine-tuning on captured data. Real-world test further confirms its resilience to potential pose errors. Our code and data will be made publicly available. Project page: https://gsfix3d.github.io.