Research map / 3D vision / 3D generation
Revisiting Monocular SLAM with Spatio-Temporal Scene Modeling
Valter Piedade, Lalit Manam, Masashi Yamazaki, Pedro Miraldo
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 |
|---|---|---|---|
| increase temporal buffer from 1 to 100 frames (ablation) | SLAM-MER · TUM room | ATE (m): 0.304 -> 0.191 | Improved |
| increase temporal buffer from 1 to 100 frames (ablation) | SLAM-MER · TUM room | FPS: 161.4 -> 98.4 | Hurt |
| increase temporal buffer to 1000 frames (ablation) | SLAM-MER · TUM room | ATE (m): 0.191 -> 0.151 (FPS 98.4 -> 45.9) | Improved |
| use smaller spatial cells (8 cm vs 32 cm) (ablation) | SLAM-MER · TUM room | FPS: 97.1 -> 30.9 | Hurt |
| vary spatial cell size (ablation) | SLAM-MER · TUM desk | ATE (m): 0.035 / 0.038 / 0.039 (8/16/32 cm) | No clear effect |
| add spatial query to temporal buffer (|B|=100) (ablation) | SLAM-MER · TUM room | ATE (m): 0.191 -> 0.163 | Improved |
| add loop closure (ablation) | SLAM-MER · TUM room | ATE (m): 0.163 -> 0.094 | Improved |
| add loop closure (ablation) | SLAM-MER · TUM desk | ATE (m): 0.034 -> 0.033 | No clear effect |
| use SLAM-MER instead of MASt3R-SLAM (uncalibrated) | SLAM-MER · TUM RGB-D / 7-Scenes | FPS: 13.2 -> 86.6 (TUM); 15.0 -> 103.2 (7-Scenes) | Improved |
| SLAM-MER vs VGGT-SLAM (SL(4)) accuracy | SLAM-MER · 7-Scenes | ATE (m): 0.059 vs 0.056 | No clear effect |
| anchor-based semi-dense map vs dense baselines | SLAM-MER · 7-Scenes dense reconstruction | Accuracy / Completion / Chamfer: Acc 0.034 vs 0.052; Comp 0.135 vs 0.058 | Mixed |
Closest papers by meaning
- SCE-SLAM: Scale-Consistent Monocular SLAM via Scene Coordinate Embeddings (CVPR 2026 · similarity 0.91)
- MASt3R-SLAM: Real-Time Dense SLAM with 3D Reconstruction Priors (CVPR 2025 · similarity 0.90)
- Photo-SLAM: Real-time Simultaneous Localization and Photorealistic Mapping for Monocular Stereo and RGB-D Cameras (CVPR 2024 · similarity 0.89)
- VGGT-SLAM: Dense RGB SLAM Optimized on the SL(4) Manifold (NeurIPS 2025 · similarity 0.88)
- GS-SLAM: Dense Visual SLAM with 3D Gaussian Splatting (CVPR 2024 · similarity 0.87)
- DG-SLAM: Robust Dynamic Gaussian Splatting SLAM with Hybrid Pose Optimization (NeurIPS 2024 · similarity 0.87)
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.
