Research map / 3D vision / Gaussian splatting
2D Gaussian Splatting-based Sparse-view Transparent Object Depth Reconstruction via Physics Simulation for Scene Update
Jeongyun Kim, Seunghoon Jeong, Giseop Kim, Myung-Hwan Jeon, Eunji Jun, Ayoung Kim
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 object-aware loss (ablation) | TRAN-D · synthetic depth, t=0 | MAE: 0.0447 -> 0.0419 | Improved |
| add object-aware loss (ablation) | TRAN-D · synthetic depth, t=0 | Gaussian count: 35983 -> 33482 | Improved |
| add object-aware loss (ablation) | TRAN-D · synthetic depth, t=1 | MAE: 0.0932 -> 0.0886 | Improved |
| add physics simulation for scene update (ablation) | TRAN-D · synthetic depth, t=1 | MAE: 0.0891 -> 0.0886 | Improved |
| add physics simulation for scene update (ablation) | TRAN-D · synthetic depth, t=1 | RMSE: 0.1945 -> 0.1936 | Improved |
| change from 3 to 12 training views (ablation) | TRAN-D · synthetic depth, t=0 | MAE: 0.0405 -> 0.0448 | No clear effect |
| change from 3 to 12 training views (ablation) | InstantSplat · synthetic depth, t=0 | MAE: 0.1306 -> 0.2062 | Hurt |
| change from 3 to 12 training views (ablation) | FSGS · synthetic depth, t=0 | MAE: 0.1846 -> 0.1426 | Improved |
| use TRAN-D object-only reconstruction vs best baseline | TRAN-D vs TranSplat · TRansPose synthetic depth, t=0 | MAE: 0.0632 -> 0.0380 | Improved |
| use TRAN-D vs 2DGS after object removal | TRAN-D vs 2DGS · TRansPose synthetic depth, t=1 | MAE: 0.0849 -> 0.0864 | Mixed |
| use TRAN-D with single bird-eye view for post-change update vs 6-image baselines | TRAN-D · efficiency, t=1 | total training time (s): 417.0 -> 13.8 | Improved |
| use TRAN-D object-only Gaussians vs 3DGS | TRAN-D · efficiency, t=0 | Gaussian count: 175.2k -> 33.5k | Improved |
Closest papers by meaning
- TranSplat: Generalizable 3D Gaussian Splatting from Sparse Multi-View Images with Transformers (AAAI 2025 · similarity 0.91)
- Reflective Gaussian Splatting (ICLR 2025 · similarity 0.90)
- Fix False Transparency by Noise Guided Splatting (NeurIPS 2025 · similarity 0.90)
- DepthSplat: Connecting Gaussian Splatting and Depth (CVPR 2025 · similarity 0.89)
- TransmissiveGS: Residual-Guided Disentangled Gaussian Splatting for Transmissive Scene Reconstruction and Rendering (NeurIPS 2026 · similarity 0.89)
- Subsurface Scattering for Gaussian Splatting (NeurIPS 2024 · 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.
