Research map / 3D vision / Gaussian splatting
X2-Gaussian: 4D Radiative Gaussian Splatting for Continuous-time Tomographic Reconstruction
Weihao Yu, Yuanhao Cai, Ruyi Zha, Zhiwen Fan, Chenxin Li, Yixuan Yuan
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 dynamic Gaussian motion modeling (DGMM) to static radiative GS (ablation) | R2-GS baseline · DIR | PSNR: 37.09 -> 38.56 | Improved |
| add self-supervised respiratory motion learning (SSRML) on top of DGMM (ablation) | X2-Gaussian · DIR | PSNR: 38.56 -> 39.34 | Improved |
| add log-space parameterization of the period (ablation) | X2-Gaussian · DIR | period estimation error (ms): 12.0 -> 5.2 | Improved |
| add bounded cycle shifts n in {-1,1} (ablation) | X2-Gaussian · DIR | period estimation error (ms): 216.8 -> 5.2 | Improved |
| add both period optimization designs (ablation) | X2-Gaussian · DIR | period estimation error (ms): 914.0 -> 5.2 | Improved |
| add bounded cycle shifts (ablation) | X2-Gaussian · DIR | PSNR: 39.28 -> 39.34 | Improved |
| add log-space parameterization (ablation) | X2-Gaussian · DIR | PSNR: 39.32 -> 39.34 | No clear effect |
| increase periodic consistency weight alpha from 0.1 to 1.0 (ablation) | X2-Gaussian · DIR | PSNR: 38.86 -> 39.34 | Improved |
| increase periodic consistency weight alpha from 1.0 to 2.0 (ablation) | X2-Gaussian · DIR | PSNR: 39.34 -> 38.41 | Hurt |
| use continuous-time Gaussian splatting with only 100 projections vs R2-GS with 300 | X2-Gaussian vs R2-GS · DIR | PSNR: 37.41 dB vs. 37.09 dB | Improved |
| replace R2-GS with X2-Gaussian | R2-GS · DIR | PSNR: 37.09 -> 39.34 | Improved |
| replace R2-GS with X2-Gaussian | R2-GS · SPARE | PSNR: 31.12 -> 32.24 | Improved |
Closest papers by meaning
- R$^2$-Gaussian: Rectifying Radiative Gaussian Splatting for Tomographic Reconstruction (NeurIPS 2024 · similarity 0.89)
- 4D Gaussian Splatting for Real-Time Dynamic Scene Rendering (CVPR 2024 · similarity 0.87)
- Gaussian-Flow: 4D Reconstruction with Dynamic 3D Gaussian Particle (CVPR 2024 · similarity 0.86)
- Fully Explicit Dynamic Gaussian Splatting (NeurIPS 2024 · similarity 0.85)
- 4D Gaussian Splatting in the Wild with Uncertainty-Aware Regularization (NeurIPS 2024 · similarity 0.85)
- DDGS-CT: Direction-Disentangled Gaussian Splatting for Realistic Volume Rendering (NeurIPS 2024 · similarity 0.85)
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.
