Research map / 3D vision / Gaussian splatting
CAG-GS: Consistent Anchor Guided Gaussian Splatting for Large-scale Scene Rendering
Shijie Xu, Qiulei Dong
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 SAM2 semantic feature alignment loss (ablation) | CAG-GS · Building | PSNR: 22.68 -> 23.46 | Improved |
| add SAM2 semantic feature alignment loss (ablation) | CAG-GS · Residence | PSNR: 21.86 -> 22.62 | Improved |
| add geometric constraints (flattening and normal consistency) (ablation) | CAG-GS · Building | PSNR: 23.18 -> 23.46 | Improved |
| add geometric constraints (ablation) | CAG-GS · Residence | LPIPS: 0.201 -> 0.202 | No clear effect |
| initialize local blocks from a coarse global model (ablation) | CAG-GS · Building | PSNR: 22.89 -> 23.46 | Improved |
| use SAM2-L instead of DINOv2-L for semantic guidance (ablation) | CAG-GS · Building | PSNR: 23.03 -> 23.46 | Improved |
| use SAM2-L instead of Depth Anything V2-L (ablation) | CAG-GS · Building | PSNR: 22.82 -> 23.46 | Improved |
| use larger SAM2 (S to L) (ablation) | CAG-GS · Building | PSNR: 23.26 -> 23.46 | Improved |
| use SAM2-B vs SAM2-L (ablation) | CAG-GS · Residence | LPIPS: 0.202 -> 0.202 | No clear effect |
| CAG-GS versus Momentum-GS | large-scale 3DGS · Rubble | LPIPS: 0.201 -> 0.221 | Hurt |
| CAG-GS versus DOGS | large-scale 3DGS · SmallCity | PSNR: 28.58 -> 27.79 | Hurt |
| semantic anchors versus explicit-representation methods | large-scale 3DGS · five large-scale scenes | SSIM / LPIPS | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 2 on most shared results.
- Momentum-GS — Momentum-GS: Momentum Gaussian Self-Distillation for High-Quality Large Scene Reconstruction (ICCV 2025) · better on 5 of 11 shared results
- DOGS — DOGS: Distributed-Oriented Gaussian Splatting for Large-Scale 3D Reconstruction Via Gaussian Consensus (NeurIPS 2024) · better on 0 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
- Speedy-Splat: Fast 3D Gaussian Splatting with Sparse Pixels and Sparse Primitives (CVPR 2025 · similarity 0.91)
- SceneSplat++: A Large Dataset and Comprehensive Benchmark for Language Gaussian Splatting (NeurIPS 2025 · similarity 0.91)
- FHGS: Feature-Homogenized Gaussian Splatting (NeurIPS 2025 · similarity 0.91)
- FlexGS: Train Once, Deploy Everywhere with Many-in-One Flexible 3D Gaussian Splatting (CVPR 2025 · similarity 0.91)
- PUP 3D-GS: Principled Uncertainty Pruning for 3D Gaussian Splatting (CVPR 2025 · similarity 0.91)
- GaussianSpa: An "Optimizing-Sparsifying" Simplification Framework for Compact and High-Quality 3D Gaussian Splatting (CVPR 2025 · similarity 0.91)
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.
