Research map / 3D vision / Gaussian splatting
GS^2: Graph-based Spatial Distribution Optimization for Compact 3D Gaussian Splatting
Xianben Yang, Tao Wang, Yuxuan Li, Yi Jin, Haibin Ling
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add GSDO (graph-based spatial distribution optimization) after pruning | LightGaussian · Mip-NeRF 360 (all 9 scenes) | PSNR: 27.35 -> 27.75 | Improved |
| add GSDO after pruning | MaskGaussian · Tanks & Temples (21 scenes) | PSNR: 24.16 -> 24.63 | Improved |
| add GSDO after pruning | LightGaussian · Tanks & Temples (21 scenes) | PSNR: 23.16 -> 23.97 | Improved |
| add opacity-aware pruning to ELBO densification (ablation) | 3DGS · Mip-NeRF 360 (all 9 scenes) | PSNR: 27.73 -> 26.79 | Hurt |
| add ELBO adaptive densification (ablation) | 3DGS · Mip-NeRF 360 (all 9 scenes) | Number of Gaussians (M): 3.12 -> 2.82 | Improved |
| add ELBO adaptive densification (ablation) | 3DGS · Tanks & Temples (21 scenes) | PSNR: 24.19 -> 24.40 | Improved |
| increase training iterations after pruning (ablation) | GS^2 pruned model · Mip-NeRF 360 (all 9 scenes) | PSNR: 26.79 -> 26.98 | No clear effect |
| add local smoothness loss and spatial optimization (Lsmt) (ablation) | pruned 3DGS · Mip-NeRF 360 (all 9 scenes) | PSNR: 26.98 -> 27.53 | Improved |
| add full GSDO (Lcet and graph encoder) on top of Lsmt (ablation) | pruned 3DGS · Tanks & Temples (21 scenes) | PSNR: 24.49 -> 24.90 | Improved |
| increase pruning interval from 100 to 400 iterations (ablation) | GS^2 · Mip-NeRF 360 (all 9 scenes) | PSNR: 27.74 -> 27.60 | No clear effect |
| add GSDO module to ADP (ablation) | GS^2 · Tanks & Temples (21 scenes) | Train time (min): 10 -> 16 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 on most shared results.
- Mini-Splatting — Mini-Splatting: Representing Scenes with a Constrained Number of Gaussians (ECCV 2024) · better on 4 of 5 shared results
- MaskGaussian — MaskGaussian: Adaptive 3D Gaussian Representation from Probabilistic Masks (CVPR 2025) · better on 4 of 4 shared results
- LightGaussian — LightGaussian: Unbounded 3D Gaussian Compression with 15x Reduction and 200+ FPS (NeurIPS 2024) · better on 4 of 4 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Speedy-Splat: Fast 3D Gaussian Splatting with Sparse Pixels and Sparse Primitives (CVPR 2025 · similarity 0.94)
- Steepest Descent Density Control for Compact 3D Gaussian Splatting (CVPR 2025 · similarity 0.92)
- Compressed 3D Gaussian Splatting for Accelerated Novel View Synthesis (CVPR 2024 · similarity 0.92)
- PUP 3D-GS: Principled Uncertainty Pruning for 3D Gaussian Splatting (CVPR 2025 · similarity 0.92)
- Gaussian Herding across Pens: An Optimal Transport Perspective on Global Gaussian Reduction for 3DGS (NeurIPS 2025 · similarity 0.92)
- GaussianSpa: An "Optimizing-Sparsifying" Simplification Framework for Compact and High-Quality 3D Gaussian Splatting (CVPR 2025 · similarity 0.92)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
