Research map / 3D vision / Gaussian splatting
MaGS: Reconstructing and Simulating Dynamic 3D Objects with Mesh-adsorbed Gaussian Splatting
Shaojie Ma, Yawei Luo, Wei Yang, Yi Yang
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 RMD-Net and RGD-Net (ablation) | MaGS · D-NeRF ablation (average) | PSNR: 41.14 -> 44.06 | Improved |
| add Gaussian Hover (ablation) | MaGS · D-NeRF ablation (average) | PSNR: 41.87 -> 44.06 | Improved |
| add RGD-Net (relative Gaussian deformation) (ablation) | MaGS · D-NeRF ablation (average) | PSNR: 42.98 -> 44.06 | Improved |
| add RGD-Net (ablation) | MaGS · D-NeRF ablation (average) | LPIPS: 0.0047 -> 0.0040 | Improved |
| add Gaussian Hover (ablation) | MaGS · D-NeRF ablation (average) | LPIPS: 0.0059 -> 0.0040 | Improved |
| MaGS versus SC-GS (second best) | SC-GS · D-NeRF (7 scenes) | PSNR: 43.31 -> 44.06 | Improved |
| MaGS versus SplattingAvatar | SplattingAvatar · PeopleSnapshot male-3-casual | PSNR: 36.48 -> 38.94 | Improved |
| MaGS versus Dynamic 2D Gaussians | Dynamic 2D Gaussians · DG-Mesh dataset | PSNR: 36.40 -> 40.76 | Improved |
| MaGS versus Dynamic 2D Gaussians on chamfer distance | Dynamic 2D Gaussians · DG-Mesh dataset | Chamfer distance: 0.5254 -> 0.6662 | Hurt |
| use 981 mesh faces versus dense meshes | Dynamic 2D Gaussians / DG-Mesh · DG-Mesh dataset | optimization time (min): 72.7 -> 47.6 | Improved |
| MaGS versus Grid4D | Grid4D · D-NeRF Lego (corrected) | PSNR: 33.24 -> 34.56 | Improved |
Later papers that compared with this method
1 later accepted paper reports MA-GS as a baseline in their results tables; 0 of them report a better number on most of the results they share with it.
- Gaussian Splatting Lucas-Kanade (ICLR 2025 · better on 0 of 1 shared result)
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 4 of 4 on most shared results.
- SC-GS — SC-GS: Sparse-Controlled Gaussian Splatting for Editable Dynamic Scenes (CVPR 2024) · better on 4 of 4 shared results
- Grid4D — Grid4D: 4D Decomposed Hash Encoding for High-Fidelity Dynamic Gaussian Splatting (NeurIPS 2024) · better on 4 of 4 shared results
- SplattingAvatar — SplattingAvatar: Realistic Real-Time Human Avatars with Mesh-Embedded Gaussian Splatting (CVPR 2024) · better on 1 of 1 shared result
- D-MiSo — D-MiSo: Editing Dynamic 3D Scenes using Multi-Gaussians Soup (NeurIPS 2024) · better on 1 of 1 shared result
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
- Motion Decoupled 3D Gaussian Splatting for Dynamic Object Representation (AAAI 2025 · similarity 0.91)
- Gaussian-Flow: 4D Reconstruction with Dynamic 3D Gaussian Particle (CVPR 2024 · similarity 0.90)
- Fully Explicit Dynamic Gaussian Splatting (NeurIPS 2024 · similarity 0.90)
- H3D-DGS: Exploring Heterogeneous 3D Motion Representation for Deformable 3D Gaussian Splatting (NeurIPS 2025 · similarity 0.90)
- SuGaR: Surface-Aligned Gaussian Splatting for Efficient 3D Mesh Reconstruction and High-Quality Mesh Rendering (CVPR 2024 · similarity 0.89)
- DGS-LRM: Real-Time Deformable 3D Gaussian Reconstruction From Monocular Videos (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.
