GSD: View-Guided Gaussian Splatting Diffusion for 3D Reconstruction

TL;DR

GSD introduces a view-guided diffusion approach utilizing Gaussian Splatting for high-quality 3D object reconstruction from a single view, improving geometry consistency and rendering quality without additional fine-tuning.

Contribution

The paper proposes a novel diffusion model based on Gaussian Splatting for 3D reconstruction, enabling view-guided generation with enhanced fidelity and explicit 3D structure from a single image.

Findings

Outperforms existing methods on CO3D dataset

Produces high-quality 3D reconstructions with detailed textures

Efficiently renders in arbitrary views

Abstract

We present GSD, a diffusion model approach based on Gaussian Splatting (GS) representation for 3D object reconstruction from a single view. Prior works suffer from inconsistent 3D geometry or mediocre rendering quality due to improper representations. We take a step towards resolving these shortcomings by utilizing the recent state-of-the-art 3D explicit representation, Gaussian Splatting, and an unconditional diffusion model. This model learns to generate 3D objects represented by sets of GS ellipsoids. With these strong generative 3D priors, though learning unconditionally, the diffusion model is ready for view-guided reconstruction without further model fine-tuning. This is achieved by propagating fine-grained 2D features through the efficient yet flexible splatting function and the guided denoising sampling process. In addition, a 2D diffusion model is further employed to enhance rendering fidelity, and improve reconstructed GS quality by polishing and re-using the rendered images. The final reconstructed objects explicitly come with high-quality 3D structure and texture, and can be efficiently rendered in arbitrary views. Experiments on the challenging real-world CO3D dataset demonstrate the superiority of our approach. Project page: https://yxmu.foo/GSD/