Research map / 3D vision / Gaussian splatting
Distractor-free Generalizable 3D Gaussian Splatting
Yanqi Bao, Jing Liao, Jing Huo, Yang Gao
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 Robust Mask to Mvsplat baseline (ablation) | Mvsplat · RobustNeRF mean (five scenes) | PSNR: 15.45 -> 17.11 | Improved |
| add reference-based mask prediction (ablation) | Mvsplat + Robust Mask · RobustNeRF mean (five scenes) | PSNR: 17.11 -> 20.35 | Improved |
| add mask refinement (ablation) | DGGS-TR · RobustNeRF mean (five scenes) | PSNR: 20.35 -> 21.02 | Improved |
| add reference entity segmentation (ablation) | DGGS-TR · RobustNeRF mean (five scenes) | PSNR: 20.79 -> 21.02 | Improved |
| add auxiliary loss (ablation) | DGGS-TR · RobustNeRF mean (five scenes) | PSNR: 20.64 -> 21.02 | Improved |
| add reference scoring mechanism at inference (ablation) | DGGS-TR · RobustNeRF mean (five scenes) | PSNR: 21.02 -> 21.47 | Improved |
| add distractor pruning at inference (ablation) | DGGS-TR + reference scoring · RobustNeRF mean (five scenes) | PSNR: 21.47 -> 21.74 | Improved |
| add pretrained segmentation model (ablation) | DGGS-TR · RobustNeRF mean (five scenes) | PSNR: 20.67 -> 21.02 | Improved |
| replace entity segmentation with SAM2 (ablation) | DGGS · RobustNeRF mean (five scenes) | PSNR: 21.74 -> 21.77 | No clear effect |
| use intersection mask fusion instead of 50% agreement (ablation) | DGGS-TR · RobustNeRF mean (five scenes) | PSNR: 20.70 -> 21.02 | Improved |
| use intersection mask fusion instead of view-angle weighting (ablation) | DGGS · RobustNeRF mean (five scenes) | PSNR: 21.60 -> 21.74 | Improved |
| use DGGS fine-tuning strategy instead of SLS-FT | Mvsplat* · Mountain fine-tuned | PSNR: 22.03 -> 23.00 | Improved |
Closest papers by meaning
- LP-3DGS: Learning to Prune 3D Gaussian Splatting (NeurIPS 2024 · similarity 0.89)
- DeSplat: Decomposed Gaussian Splatting for Distractor-Free Rendering (CVPR 2025 · similarity 0.89)
- Robust Neural Rendering in the Wild with Asymmetric Dual 3D Gaussian Splatting (NeurIPS 2025 · similarity 0.89)
- PUP 3D-GS: Principled Uncertainty Pruning for 3D Gaussian Splatting (CVPR 2025 · similarity 0.89)
- GaussianSpa: An "Optimizing-Sparsifying" Simplification Framework for Compact and High-Quality 3D Gaussian Splatting (CVPR 2025 · similarity 0.89)
- Metropolis-Hastings Sampling for 3D Gaussian Reconstruction (NeurIPS 2025 · similarity 0.89)
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.
