Generative Densification: Learning to Densify Gaussians for High-Fidelity Generalizable 3D Reconstruction

TL;DR

This paper introduces Generative Densification, a method that enhances Gaussian-based 3D reconstruction by efficiently up-sampling features to better capture high-frequency details, improving generalization and detail fidelity.

Contribution

It proposes a novel densification technique that up-samples features from feed-forward models to generate fine Gaussians in a single pass, surpassing previous iterative methods.

Findings

Outperforms state-of-the-art methods in object and scene reconstruction

Achieves better detail representation with smaller models

Demonstrates strong generalization across tasks

Abstract

Generalized feed-forward Gaussian models have achieved significant progress in sparse-view 3D reconstruction by leveraging prior knowledge from large multi-view datasets. However, these models often struggle to represent high-frequency details due to the limited number of Gaussians. While the densification strategy used in per-scene 3D Gaussian splatting (3D-GS) optimization can be adapted to the feed-forward models, it may not be ideally suited for generalized scenarios. In this paper, we propose Generative Densification, an efficient and generalizable method to densify Gaussians generated by feed-forward models. Unlike the 3D-GS densification strategy, which iteratively splits and clones raw Gaussian parameters, our method up-samples feature representations from the feed-forward models and generates their corresponding fine Gaussians in a single forward pass, leveraging the embedded prior knowledge for enhanced generalization. Experimental results on both object-level and scene-level reconstruction tasks demonstrate that our method outperforms state-of-the-art approaches with comparable or smaller model sizes, achieving notable improvements in representing fine details.