Research map / 3D vision / Gaussian splatting
RaPTGS: Render-Agnostic Post-Training Compression of 3D Gaussian Splatting
Asif Jawad, K. M. Azwad Hossain
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 Voxel Mass Compensation after pruning (ablation) | RaPTGS · Mip-NeRF 360 | PSNR: 26.57 -> 26.74 | Improved |
| add degree-wise SH quantization (ablation) | RaPTGS · Mip-NeRF 360 | size (MB) / PSNR: 437.39 -> 119.93 MB, PSNR 26.74 -> 26.40 | Mixed |
| add Hilbert sort reordering before entropy coding (ablation) | RaPTGS · Mip-NeRF 360 | size (MB): 32.66 -> 25.39 | Improved |
| add windowed 2-opt after Hilbert sort (ablation) | RaPTGS · Mip-NeRF 360 | size (MB): 25.39 -> 24.55 | Improved |
| add spatial reordering under npz archive coding (ablation) | RaPTGS (npz) · Mip-NeRF 360 | size (MB): 34.91 -> 29.98 | Improved |
| use max pooling instead of sum pooling of proxies (ablation) | RaPTGS importance score · Mip-NeRF 360 | PSNR: 26.48 -> 26.57 | Improved |
| add opacity salience proxy to the importance score (ablation) | RaPTGS importance score · Mip-NeRF 360 | PSNR: 26.54 -> 26.57 | Improved |
| add shape anisotropy proxy to the importance score (ablation) | RaPTGS importance score · Mip-NeRF 360 | PSNR: 26.53 -> 26.57 | Improved |
| add scale-aware appearance energy proxy (ablation) | RaPTGS importance score · Mip-NeRF 360 | PSNR: 25.98 -> 26.57 | Improved |
| add lower bound on extent in SAE (ablation) | RaPTGS SAE proxy · Mip-NeRF 360 | PSNR: 26.46 -> 26.57 | Improved |
| use VMC at prune ratios above 0.4 versus 0.0 - 0.3 (ablation) | RaPTGS · Mip-NeRF 360 | PSNR: improves above 0.4, marginal at 0.0 - 0.3 | Mixed |
| replace archive coding with JPEG XL (ablation) | RaPTGS · Mip-NeRF 360 | rate-distortion | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 2 on most shared results.
- MesonGS — MesonGS: Post-training Compression of 3D Gaussians via Efficient Attribute Transformation (ECCV 2024) · better on 6 of 7 shared results
- PUP 3D-GS — PUP 3D-GS: Principled Uncertainty Pruning for 3D Gaussian Splatting (CVPR 2025) · better on 1 of 3 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
- Fast Feedforward 3D Gaussian Splatting Compression (ICLR 2025 · similarity 0.94)
- Compressed 3D Gaussian Splatting for Accelerated Novel View Synthesis (CVPR 2024 · similarity 0.94)
- Optimized Minimal 3D Gaussian Splatting (NeurIPS 2025 · similarity 0.93)
- ContextGS : Compact 3D Gaussian Splatting with Anchor Level Context Model (NeurIPS 2024 · similarity 0.93)
- LP-3DGS: Learning to Prune 3D Gaussian Splatting (NeurIPS 2024 · similarity 0.93)
- Locality-aware Gaussian Compression for Fast and High-quality Rendering (ICLR 2025 · similarity 0.93)
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.
