Research map / 3D vision / Gaussian splatting
Faster and Better 3D Splatting via Group Training
Chengbo Wang, Guozheng Ma, Yifei Xue, Yizhen Lao
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 opacity-based prioritized sampling (OPS) instead of random sampling to Group Training (ablation) | 3DGS · Mip-NeRF360 | training time (min): 22.58 -> 19.56 | Improved |
| add OPS instead of random sampling (ablation) | 3DGS · Mip-NeRF360 | model size (MB): 902 -> 679 | Improved |
| add Group Training (OPS, 0~30K) to 3DGS | 3DGS · Mip-NeRF360 | training time (min): 26.72 -> 19.56 | Improved |
| add Group Training (OPS, 0~30K) to 3DGS | 3DGS · Tanks & Temples | training time (min): 15.00 -> 10.95 | Improved |
| add Group Training (OPS, 0~30K) to 3DGS | 3DGS · Deep Blending | training time (min): 23.90 -> 16.85 | Improved |
| add Group Training (OPS, 0~30K) to 3DGS | 3DGS · NeRF-Synthetic | training time (min): 6.06 -> 4.79 | Improved |
| add Group Training (OPS, 0~30K) to 3DGS | 3DGS · Deep Blending | PSNR: 29.586 -> 29.746 | Improved |
| add Group Training with random sampling (0~15K) to 3DGS | 3DGS · Deep Blending | PSNR: 29.586 -> 29.185 | Hurt |
| add Group Training (OPS, 0~30K) to Mip-Splatting | Mip-Splatting · Deep Blending | training time (min): 35.1 -> 24.0 | Improved |
| add Group Training (OPS, 0~30K) to Mip-Splatting | Mip-Splatting · Tanks & Temples | PSNR: 23.749 -> 24.156 | Improved |
| add Group Training with random sampling (0~15K) to Mip-Splatting | Mip-Splatting · Tanks & Temples | training time (min): 23.0 -> 24.6 | Hurt |
| add global densification to cyclic resampling (ablation) | Group Training w/ OPS · Tanks & Temples | model size (MB): 292 -> 231 | Improved |
Closest papers by meaning
- Speedy-Splat: Fast 3D Gaussian Splatting with Sparse Pixels and Sparse Primitives (CVPR 2025 · similarity 0.92)
- DOGS: Distributed-Oriented Gaussian Splatting for Large-Scale 3D Reconstruction Via Gaussian Consensus (NeurIPS 2024 · similarity 0.92)
- GaussianPro: 3D Gaussian Splatting with Progressive Propagation (ICML 2024 · similarity 0.92)
- LP-3DGS: Learning to Prune 3D Gaussian Splatting (NeurIPS 2024 · similarity 0.92)
- SfM-Free 3D Gaussian Splatting via Hierarchical Training (CVPR 2025 · similarity 0.92)
- FlexGS: Train Once, Deploy Everywhere with Many-in-One Flexible 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.
