Research map / 3D vision / Gaussian splatting
Per-Gaussian Embedding-Based Deformation for Deformable 3D Gaussian Splatting
Jeongmin Bae, Seoha Kim, Youngsik Yun, Hahyun Lee, Gun Bang, Youngjung Uh
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 fine deformation to coarse-only deformation (ablation) | E-D3DGS · Neural 3D Video | PSNR: 29.48 -> 29.70 | Improved |
| add coarse deformation to fine-only deformation (ablation) | E-D3DGS · Neural 3D Video | PSNR: 29.23 -> 29.70 | Improved |
| inject Gaussian center coordinates into the decoders (ablation) | E-D3DGS · Neural 3D Video | PSNR: 29.70 -> 29.60 | Hurt |
| add local smoothness regularization on per-Gaussian embeddings (ablation) | E-D3DGS · regularization ablation | PSNR: 32.26 -> 32.34 | Improved |
| use physically-based regularization instead of embedding smoothness regularization (ablation) | E-D3DGS · regularization ablation | PSNR: 32.34 -> 32.08 | Hurt |
| increase embedding sizes (temporal 128-512, Gaussian 16-64) (ablation) | E-D3DGS · flame_steak | PSNR: 33.53 - 33.71 | No clear effect |
| use per-Gaussian embedding deformation instead of coordinate-based field (4DGaussians) | deformable 3DGS · Neural 3D Video | PSNR: 30.71 -> 31.31 | Improved |
| use per-Gaussian embedding deformation instead of 4DGaussians | deformable 3DGS · Technicolor | PSNR: 29.62 -> 33.24 | Improved |
| use per-Gaussian embedding deformation instead of 4DGaussians | deformable 3DGS · Neural 3D Video | training time: 50 mins -> 1 hours 52 mins | Hurt |
| use E-D3DGS instead of HyperNeRF DS on HyperNeRF | deformable 3DGS · HyperNeRF | LPIPS: 0.153 -> 0.231 | Hurt |
Closest papers by meaning
- MotionGS: Exploring Explicit Motion Guidance for Deformable 3D Gaussian Splatting (NeurIPS 2024 · similarity 0.92)
- Motion Decoupled 3D Gaussian Splatting for Dynamic Object Representation (AAAI 2025 · similarity 0.92)
- Deformable 3D Gaussians for High-Fidelity Monocular Dynamic Scene Reconstruction (CVPR 2024 · similarity 0.91)
- H3D-DGS: Exploring Heterogeneous 3D Motion Representation for Deformable 3D Gaussian Splatting (NeurIPS 2025 · similarity 0.91)
- Gaussian-Flow: 4D Reconstruction with Dynamic 3D Gaussian Particle (CVPR 2024 · similarity 0.91)
- DGS-LRM: Real-Time Deformable 3D Gaussian Reconstruction From Monocular Videos (NeurIPS 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.
