Search papers, labs, and topics across Lattice.
This paper introduces GS^2, a novel approach to compact 3D Gaussian Splatting that optimizes the spatial distribution of Gaussian points to improve rendering quality and reduce memory footprint. GS^2 employs an ELBO-based adaptive densification strategy for controlled point addition, opacity-aware progressive pruning for dynamic point removal, and a graph-based feature encoding module for feature-guided point shifting. Experiments demonstrate that GS^2 achieves higher PSNR with significantly fewer Gaussian points (12.5% of original 3DGS) and outperforms existing methods in both rendering quality and memory efficiency.
Achieve comparable or better novel view synthesis with 8x fewer Gaussians by intelligently redistributing points in space.
3D Gaussian Splatting (3DGS) has demonstrated breakthrough performance in novel view synthesis and real-time rendering. Nevertheless, its practicality is constrained by the high memory cost due to a huge number of Gaussian points. Many pruning-based 3DGS variants have been proposed for memory saving, but often compromise spatial consistency and may lead to rendering artifacts. To address this issue, we propose graph-based spatial distribution optimization for compact 3D Gaussian Splatting (GS\textasciicircum2), which enhances reconstruction quality by optimizing the spatial distribution of Gaussian points. Specifically, we introduce an evidence lower bound (ELBO)-based adaptive densification strategy that automatically controls the densification process. In addition, an opacity-aware progressive pruning strategy is proposed to further reduce memory consumption by dynamically removing low-opacity Gaussian points. Furthermore, we propose a graph-based feature encoding module to adjust the spatial distribution via feature-guided point shifting. Extensive experiments validate that GS\textasciicircum2 achieves a compact Gaussian representation while delivering superior rendering quality. Compared with 3DGS, it achieves higher PSNR with only about 12.5\% Gaussian points. Furthermore, it outperforms all compared baselines in both rendering quality and memory efficiency.
