Research map / 3D vision / Gaussian splatting
A Constrained Optimization Approach for Gaussian Splatting from Coarsely-posed Images and Noisy Lidar Point Clouds
Jizong Peng, Tze Ho Elden Tse, Kai Xu, Wenchao Gao, Angela Yao
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 camera decomposition (without pre-conditioning) (ablation) | Ours · Cafeteria | SSIM: 0.8659 -> 0.9154 | Improved |
| add sensitivity-based pre-conditioning on top of camera decomposition (ablation) | Ours · Cafeteria | SSIM: 0.9154 -> 0.9168 | Improved |
| add pre-conditioning alone, without camera decomposition (ablation) | Ours · Cafeteria | SSIM: 0.8659 -> 0.8577 | Hurt |
| add intrinsic refinement (ablation) | Ours · Cafeteria | PSNR: 27.40 -> 29.05 | Improved |
| add intrinsic refinement (ablation) | Ours · Laboratory | PSNR: 26.79 -> 28.64 | Improved |
| add log-barrier constraint of +-2% feasible region (ablation) | Ours · Cafeteria | SSIM: 6.8% | Improved |
| add reprojection error regularization (ablation) | Ours · Cafeteria | SSIM: 0.8945 -> 0.9130 | Improved |
| add soft epipolar constraint (ablation) | Ours · Cafeteria | SSIM: 0.8945 -> 0.9141 | Improved |
| add both epipolar and reprojection constraints at 0.5 degree pose noise (ablation) | Ours · Cafeteria (0.5 degree noise) | PSNR: 24.80 -> 25.20 | Improved |
| add test-time pose refinement (ablation) | Ours · Cafeteria | PSNR: 19.80 -> 23.04 | Improved |
| add exposure compensation alone (ablation) | Ours · Cafeteria | SSIM: 0.7752 -> 0.7872 | No clear effect |
| add test-time pose refinement and exposure compensation together (ablation) | Ours · Cafeteria | PSNR: 19.80 -> 28.58 | Improved |
Closest papers by meaning
- KeyGS: A Keyframe-Centric Gaussian Splatting Method for Monocular Image Sequences (AAAI 2025 · similarity 0.91)
- EAP-GS: Efficient Augmentation of Pointcloud for 3D Gaussian Splatting in Few-shot Scene Reconstruction (CVPR 2025 · similarity 0.91)
- SfM-Free 3D Gaussian Splatting via Hierarchical Training (CVPR 2025 · similarity 0.90)
- COLMAP-Free 3D Gaussian Splatting (CVPR 2024 · similarity 0.90)
- SplatFormer: Point Transformer for Robust 3D Gaussian Splatting (ICLR 2025 · similarity 0.90)
- MotionGS: Exploring Explicit Motion Guidance for Deformable 3D Gaussian Splatting (NeurIPS 2024 · similarity 0.90)
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.
