Research map / 3D vision / Gaussian splatting
AnchorSplat: Fast and Structure Consistent Detail Synthesis for Gaussian Splatting
Dexu Zhu, Jiangnan Shao, Xiaofeng Wang, Junxian Duan, Jie Cao, Zheng Zhu, Huaibo Huang
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 Point Anchor Mechanism (local offset constraint on generated primitives) (ablation) | AnchorSplat · 3DGS-SR | PSNR: 26.79 -> 36.57 | Improved |
| add Point Anchor Mechanism (ablation) | AnchorSplat · 3DGS-SR | LPIPS: 0.186 -> 0.058 | Improved |
| increase multiplicative primitive factor K from 1 to 20 (ablation) | AnchorSplat · 3DGS-SR | PSNR: 36.42 -> 36.57 | Improved |
| increase K from 1 to 20 (ablation) | AnchorSplat · 3DGS-SR | LPIPS: 0.063 -> 0.058 | Improved |
| increase K from 10 to 20 (ablation) | AnchorSplat · 3DGS-SR | PSNR: 36.51 -> 36.57 | Improved |
| AnchorSplat feed-forward vs best 2D-centric optimization baseline | AnchorSplat · 3DGS-SR | PSNR: 35.69 -> 36.57 | Improved |
| AnchorSplat vs optimization methods in processing time | AnchorSplat · 3DGS-SR | processing time: ~ 34m -> ~ 0.01s (vs Sequence Matters) | Improved |
| apply AnchorSplat zero-shot to low-quality 3DGS assets | 3DGS asset · NeRF-synthetic (zero-shot) | PSNR: 23.30 -> 28.97 | Improved |
| AnchorSplat (source-free) vs source-dependent methods zero-shot | AnchorSplat · NeRF-synthetic (zero-shot) | PSNR: 28.97 vs 31.41 (Sequence Matters), 30.83 (SRGS); vs 28.44 (SuperGaussian) | Mixed |
| AnchorSplat vs source-dependent methods zero-shot on SSIM/LPIPS | AnchorSplat · NeRF-synthetic (zero-shot) | SSIM: 0.935 vs 0.945 (SuperGaussian) | Hurt |
Closest papers by meaning
- AnchorSplat: Feed-Forward 3D Gaussian Splatting With 3D Geometric Priors (CVPR 2026 · similarity 0.94)
- SplatFormer: Point Transformer for Robust 3D Gaussian Splatting (ICLR 2025 · similarity 0.92)
- Speedy-Splat: Fast 3D Gaussian Splatting with Sparse Pixels and Sparse Primitives (CVPR 2025 · similarity 0.92)
- ReorgGS: Equivalent Distribution Reorganization for 3D Gaussian Splatting (NeurIPS 2026 · similarity 0.92)
- GaussianSpa: An "Optimizing-Sparsifying" Simplification Framework for Compact and High-Quality 3D Gaussian Splatting (CVPR 2025 · similarity 0.92)
- GaussianPro: 3D Gaussian Splatting with Progressive Propagation (ICML 2024 · 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.
