Research map / 3D vision / Gaussian splatting
CompGS: Smaller and Faster Gaussian Splatting with Vector Quantization
K L Navaneet, Kossar Pourahmadi Meibodi, Soroush Abbasi Koohpayegani, Hamed Pirsiavash
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 K-Means vector quantization and opacity regularization to 3DGS | 3DGS · Mip-NeRF360 | memory (MB): 778 -> 19 | Improved |
| add K-Means vector quantization and opacity regularization to 3DGS | 3DGS · Mip-NeRF360 | FPS: 149 -> 344 | Improved |
| add CompGS compression | 3DGS · Mip-NeRF360 | PSNR: 27.42 -> 27.12 | Hurt |
| add CompGS compression | 3DGS · Deep Blending | PSNR: 29.49 -> 29.75 | Improved |
| quantization-aware K-Means training instead of post-training clustering (ablation) | CompGS · Mip-NeRF360 | PSNR: 25.46 -> 26.97 | Improved |
| quantize position parameters (ablation) | CompGS · Tanks&Temples Train | PSNR: 21.99 -> 19.81 | Hurt |
| exclude positions from 8-bit quantization (ablation) | 3DGS bit quantization · Mip-NeRF360 | PSNR: 14.41 -> 27.38 | Improved |
| reduce precision to 4-bit even without quantizing positions (ablation) | 3DGS bit quantization · Mip-NeRF360 | PSNR: 27.38 -> 17.42 | Hurt |
| use K-Means quantization instead of dropping spherical harmonics (ablation) | 3DGS · Mip-NeRF360 | PSNR: 26.80 -> 26.97 | Improved |
| use opacity regularization instead of tuning densification/pruning thresholds (ablation) | 3DGS Gaussian count reduction · Mip-NeRF360 | PSNR: 26.57 -> 27.42 vs grad thresh baseline | Improved |
| increase opacity regularization lambda_reg from 1.0 to 3.0 (x1e-7) (ablation) | CompGS · Mip-NeRF360 | PSNR: 27.12 -> 26.83; Gaussians 845K -> 390K | Hurt |
| update K-Means assignments every 500 iterations instead of 100 (ablation) | CompGS · Mip-NeRF360 | PSNR: 26.95 -> 26.15; time 22.5 -> 18.1 | Hurt |
Later papers that compared with this method
3 later accepted papers report CompGS as a baseline in their results tables; 2 of them report a better number on most of the results they share with it.
- Penetration-Free Compositional 3D Generation via Gaussian Surface Offset (ECCV 2026 · better on 8 of 8 shared results)
- NeuralGS: Bridging Neural Fields and 3D Gaussian Splatting for Compact 3D Representations (AAAI 2026 · better on 6 of 6 shared results)
- Locality-aware Gaussian Compression for Fast and High-quality Rendering (ICLR 2025 · better on 0 of 2 shared results)
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- LightGaussian — LightGaussian: Unbounded 3D Gaussian Compression with 15x Reduction and 200+ FPS (NeurIPS 2024) · better on 2 of 2 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
- Optimized Minimal 3D Gaussian Splatting (NeurIPS 2025 · similarity 0.94)
- Compact 3D Gaussian Representation for Radiance Field (CVPR 2024 · similarity 0.93)
- GaussianPro: 3D Gaussian Splatting with Progressive Propagation (ICML 2024 · similarity 0.93)
- 3D-HGS: 3D Half-Gaussian Splatting (CVPR 2025 · similarity 0.92)
- FlashGS: Efficient 3D Gaussian Splatting for Large-scale and High-resolution Rendering (CVPR 2025 · similarity 0.92)
- IBGS: Image-Based Gaussian Splatting (NeurIPS 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.
