Research map / 3D vision / Gaussian splatting
RGBD GS-ICP SLAM
Seongbo Ha, Jiung Yeon, Hyeonwoo Yu
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 |
|---|---|---|---|
| replace plane scale regularization with ellipse regularization in G-ICP tracking (ablation) | RGBD GS-ICP SLAM · TUM-RGBD | ATE RMSE avg (cm): 29.12 -> 2.37 | Improved |
| reuse G-ICP covariances to initialize new Gaussians instead of simple-knn scales (ablation) | RGBD GS-ICP SLAM · Replica | ATE (cm): 8.893 -> 0.258 | Improved |
| divide new Gaussian scale by z^1.5 (scale aligning) (ablation) | RGBD GS-ICP SLAM · Replica | PSNR (dB): 33.21 -> 38.83 | Improved |
| use z^1.5 instead of z for scale normalization (ablation) | RGBD GS-ICP SLAM · Replica | PSNR (dB): 38.70 -> 38.83 | No clear effect |
| add mapping-only keyframes every 10 frames separate from tracking keyframes (ablation) | RGBD GS-ICP SLAM · Replica | ATE (cm): 0.176 -> 0.157 | Improved |
| train mapping on random keyframes instead of the most recent keyframe (ablation) | RGBD GS-ICP SLAM · Replica | PSNR (dB): 26.85 -> 38.83 | Improved |
| add Gaussian densification during mapping (ablation) | RGBD GS-ICP SLAM · Replica | ATE (cm): 0.157 -> 0.188 | Hurt |
| add Gaussian pruning during mapping (ablation) | RGBD GS-ICP SLAM · Replica | ATE (cm): 0.159 -> 0.157 | No clear effect |
| limit tracking to 30 FPS instead of unlimited speed | RGBD GS-ICP SLAM · Replica | PSNR (dB): 35.93 -> 38.83 | Improved |
| use G-ICP coupled single-map tracking versus SplaTAM on TUM-RGBD | RGBD GS-ICP SLAM · TUM-RGBD | PSNR (dB): 23.46 -> 20.72 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 4 of 4 on most shared results.
- SplaTAM — SplaTAM: Splat Track & Map 3D Gaussians for Dense RGB-D SLAM (CVPR 2024) · better on 5 of 6 shared results
- GS-SLAM — GS-SLAM: Dense Visual SLAM with 3D Gaussian Splatting (CVPR 2024) · better on 2 of 2 shared results
- Photo-SLAM — Photo-SLAM: Real-time Simultaneous Localization and Photorealistic Mapping for Monocular Stereo and RGB-D Cameras (CVPR 2024) · better on 1 of 1 shared result
- NICE-SLAM — NICE-SLAM: Neural Implicit Scalable Encoding for SLAM (CVPR 2022) · better on 1 of 1 shared result
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 0.92)
- VTGaussian-SLAM: RGBD SLAM for Large Scale Scenes with Splatting View-Tied 3D Gaussians (ICML 2025 · similarity 0.90)
- Gaussian Splatting SLAM (CVPR 2024 · similarity 0.90)
- SplaTAM: Splat Track & Map 3D Gaussians for Dense RGB-D SLAM (CVPR 2024 · similarity 0.90)
- WildGS-SLAM: Monocular Gaussian Splatting SLAM in Dynamic Environments (CVPR 2025 · similarity 0.88)
- Benchmarking Implicit Neural Representation and Geometric Rendering in Real-Time RGB-D SLAM (CVPR 2024 · 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.
