Research map / 3D vision / Gaussian splatting
Manifold-Aware Spectral Compaction: A Graph Signal Processing Perspective on Online Gaussian Reduction for 3DGS SLAM
Jiayi Tian, Jiaze Wang, Tian Xia, Wenzhe zhao, Pengju Ren
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 |
|---|---|---|---|
| use spectral MASC compaction instead of random subsampling (ablation) | DROID-Splat at 40% retention · Replica (RGB-D, rho=40% ablation) | PSNR: 28.15 -> 39.16 | Improved |
| use spectral MASC compaction instead of heuristic pruning (ablation) | DROID-Splat at 40% retention · Replica (RGB-D, rho=40% ablation) | PSNR: 35.20 -> 39.16 | Improved |
| use spectral clustering instead of Euclidean K-Means (ablation) | DROID-Splat at 40% retention · Replica (RGB-D, rho=40% ablation) | ATE (cm): 0.40 -> 0.33 | Improved |
| add appearance cues to graph weights (ablation) | MASC · TUM RGB-D (rho=40%) | PSNR: 19.45 -> 20.96 | Improved |
| add SKL divergence to graph weights (versus L2) (ablation) | MASC · TUM RGB-D (rho=40%) | PSNR: 19.85 -> 20.96 | Improved |
| use Wasserstein barycenter aggregation instead of simple mean (ablation) | MASC · TUM RGB-D (rho=40%) | ATE (cm): 1.45 -> 1.42 | Improved |
| use 1000 Nystrom landmarks instead of 100 (ablation) | MASC · TUM RGB-D (rho=40%) | PSNR: 18.20 -> 20.96 | Mixed |
| add MASC compaction (40% retention) | DROID-Splat (RGB-D) · Replica | GPU memory: 14.00 GB -> 5.00 GB | Improved |
| add MASC compaction (40% retention) | DROID-Splat (RGB-D) · Replica | R-FPS: 145 -> 350 | Improved |
| add MASC compaction (40% retention) | DROID-Splat (RGB-D) · Replica | PSNR: 39.66 -> 39.16 | No clear effect |
| add MASC compaction (40% retention) | Photo-SLAM · TUM RGB-D (average) | ATE (cm): 1.32 -> 0.98 | Improved |
| add MASC compaction | Photo-SLAM (RGB-D) · Replica | ATE RMSE (cm): 0.60 -> 0.30 | Improved |
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.
- Photo-SLAM — Photo-SLAM: Real-time Simultaneous Localization and Photorealistic Mapping for Monocular Stereo and RGB-D Cameras (CVPR 2024) · better on 6 of 6 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
- GS-SLAM: Dense Visual SLAM with 3D Gaussian Splatting (CVPR 2024 · similarity 0.88)
- VTGaussian-SLAM: RGBD SLAM for Large Scale Scenes with Splatting View-Tied 3D Gaussians (ICML 2025 · similarity 0.88)
- Gaussian Herding across Pens: An Optimal Transport Perspective on Global Gaussian Reduction for 3DGS (NeurIPS 2025 · similarity 0.88)
- Gaussian Splatting SLAM (CVPR 2024 · similarity 0.88)
- DG-SLAM: Robust Dynamic Gaussian Splatting SLAM with Hybrid Pose Optimization (NeurIPS 2024 · similarity 0.88)
- VGGT-SLAM: Dense RGB SLAM Optimized on the SL(4) Manifold (NeurIPS 2025 · similarity 0.88)
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.
