Research map / 3D vision / Gaussian splatting
No Pose at All: Self-Supervised Pose-Free 3D Gaussian Splatting from Sparse Views
Ranran Huang, Krystian Mikolajczyk
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 reprojection loss (ablation) | SPFSplat · RE10K NVS* (EPA) | PSNR: 21.914 -> 25.845 | Improved |
| add reprojection loss (ablation) | SPFSplat · RE10K pose | AUC@5: 0.028 -> 0.617 | Improved |
| add intrinsic embedding (ablation) | SPFSplat · RE10K NVS* (EPA) | PSNR: 25.519 -> 25.845 | Improved |
| add intrinsic embedding (ablation) | SPFSplat · RE10K pose | AUC@5: 0.562 -> 0.617 | Improved |
| add ground-truth pose supervision loss (ablation) | SPFSplat · RE10K pose | AUC@5: 0.617 -> 0.691 | Improved |
| add ground-truth pose supervision loss (ablation) | SPFSplat · RE10K NVS* (EPA) | PSNR: 25.845 -> 25.910 | No clear effect |
| add DL3DV to training data (ablation) | SPFSplat · RE10K pose | AUC@5: 0.617 -> 0.635 | Improved |
| add DL3DV to training data (ablation) | SPFSplat · ACID pose | AUC@5: 0.364 -> 0.395 | Improved |
| increase input views from 2 to 10 (ablation) | SPFSplat · RE10K NVS | PSNR: 25.403 -> 27.159 | Improved |
| add L2 loss (ablation) | SPFSplat · RE10K NVS (predicted pose) | PSNR: 24.310 -> 25.484 | Improved |
| add LPIPS loss (ablation) | SPFSplat · RE10K NVS (predicted pose) | LPIPS: 0.210 -> 0.153 | Improved |
| initialize with MASt3R instead of random (ablation) | SPFSplat · RE10K NVS | PSNR: 21.200 -> 25.484 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 5 of 5 on most shared results.
- PF3plat — PF3plat: Pose-Free Feed-Forward 3D Gaussian Splatting for Novel View Synthesis (ICML 2025) · better on 4 of 4 shared results
- SelfSplat — SelfSplat: Pose-Free and 3D Prior-Free Generalizable 3D Gaussian Splatting (CVPR 2025) · better on 3 of 3 shared results
- pixelSplat — pixelSplat: 3D Gaussian Splats from Image Pairs for Scalable Generalizable 3D Reconstruction (CVPR 2024) · better on 3 of 3 shared results
- MVSplat — MVSplat: Efficient 3D Gaussian Splatting from Sparse Multi-View Images (ECCV 2024) · better on 2 of 2 shared results
- DUSt3R — DUSt3R: Geometric 3D Vision Made Easy (CVPR 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
- No Pose, No Problem: Surprisingly Simple 3D Gaussian Splats from Sparse Unposed Images (ICLR 2025 · similarity 0.93)
- PF3plat: Pose-Free Feed-Forward 3D Gaussian Splatting for Novel View Synthesis (ICML 2025 · similarity 0.93)
- SelfSplat: Pose-Free and 3D Prior-Free Generalizable 3D Gaussian Splatting (CVPR 2025 · similarity 0.92)
- COLMAP-Free 3D Gaussian Splatting (CVPR 2024 · similarity 0.92)
- pixelSplat: 3D Gaussian Splats from Image Pairs for Scalable Generalizable 3D Reconstruction (CVPR 2024 · similarity 0.90)
- SfM-Free 3D Gaussian Splatting via Hierarchical Training (CVPR 2025 · 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.
