Research map / 3D vision / Gaussian splatting
Learning Unified Representation of 3D Gaussian Splatting
Yuelin Xin, Yuheng Liu, Xiaohui Xie, Xinke Li
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 |
|---|---|---|---|
| encode Gaussians as submanifold fields instead of raw parameters | VAE embedding · ShapeSplat | PSNR: 37.512 -> 63.408 | Improved |
| encode Gaussians as submanifold fields instead of raw parameters | VAE embedding · Mip-NeRF 360 | PSNR: 18.818 -> 29.833 | Improved |
| use SF-VAE point-cloud decoder with parametric input (ablation) | parametric VAE · ShapeSplat | PSNR: 37.512 -> 44.725 | Improved |
| use submanifold field input in addition to the SF decoder (ablation) | VAE embedding · Mip-NeRF 360 | PSNR: 20.923 -> 29.833 | Improved |
| submanifold field representation under cross-domain transfer | SF-VAE · train ShapeSplat, test Mip-NeRF 360 | PSNR: 9.753 -> 19.194 | Improved |
| submanifold field instead of parametric (L1 distance) | SF-VAE · ShapeSplat | L1-Dist: 0.040 -> 0.098 | Hurt |
| increase sampled points P beyond 12^2 (ablation) | SF-VAE · Mip-NeRF 360 | reconstruction quality | No clear effect |
| replace PointNet encoder with DGCNN (ablation) | SF-VAE · Mip-NeRF 360 | reconstruction fidelity | No clear effect |
| GPU FP32 fitting module instead of CPU FP64 (ablation) | SF-VAE · Mip-NeRF 360 | PSNR: -0.4 PSNR, 85x speedup | No clear effect |
| use SF embedding as regression target for Gaussian Neural Field | Gaussian Neural Field · visual fidelity | rendering quality | Improved |
Closest papers by meaning
- UVGS: Reimagining Unstructured 3D Gaussian Splatting using UV Mapping (CVPR 2025 · similarity 0.91)
- Effective Rank Analysis and Regularization for Enhanced 3D Gaussian Splatting (NeurIPS 2024 · similarity 0.90)
- GaussianPro: 3D Gaussian Splatting with Progressive Propagation (ICML 2024 · similarity 0.90)
- ContextGS : Compact 3D Gaussian Splatting with Anchor Level Context Model (NeurIPS 2024 · similarity 0.90)
- 3D Student Splatting and Scooping (CVPR 2025 · similarity 0.90)
- SfM-Free 3D Gaussian Splatting via Hierarchical Training (CVPR 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.
