Research map / 3D vision / Gaussian splatting
Sparfels: Fast Reconstruction from Sparse Unposed Imagery
Shubhendu Jena, Amine Ouasfi, Mae Younes, Adnane Boukhayma
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 bundle adjustment of cameras (ablation) | 2DGS · DTU ablation (3-view, 5 scenes) | Rel: 7.93 -> 7.79 | Improved |
| add bundle adjustment of cameras (ablation) | 2DGS · DTU ablation (3-view, 5 scenes) | ATE: 0.0327 -> 0.0284 | Improved |
| add MASt3R correspondence loss (ablation) | BA-2DGS · DTU ablation (3-view, 5 scenes) | Rel: 7.79 -> 7.41 | Improved |
| add MASt3R correspondence loss (ablation) | BA-2DGS · DTU ablation (3-view, 5 scenes) | ATE: 0.0284 -> 0.0256 | Improved |
| initialize splat orientation with pointmap normals (ablation) | BA-2DGS + L_corr · DTU ablation (3-view, 5 scenes) | Rel: 7.41 -> 7.03 | Improved |
| initialize splat orientation with pointmap normals (ablation) | BA-2DGS + L_corr · DTU ablation (3-view, 5 scenes) | Normal consistency: 0.804 -> 0.829 | Improved |
| add splatted color variance loss (ablation) | BA-2DGS + L_corr + normal init · DTU ablation (3-view, 5 scenes) | Rel: 7.03 -> 6.79 | Improved |
| add splatted color variance loss (ablation) | BA-2DGS + L_corr + normal init · DTU ablation (3-view, 5 scenes) | Normal consistency: 0.829 -> 0.834 | Improved |
| add splatted color variance loss (ablation) | BA-2DGS + L_corr + normal init · DTU ablation (3-view, 5 scenes) | ATE: 0.0237 -> 0.0238 | No clear effect |
| use Sparfels instead of best baseline InstSplatGOF for camera poses | Sparfels · DTU (3-view, 15 scans) | ATE: 13.1% reduction in error | Improved |
| use Sparfels instead of InstantSplat-XL | Sparfels · MVImgNet (3-view) | ATE: 0.0184 -> 0.0273 | Hurt |
| use Sparfels instead of InstantSplat-XL | Sparfels · Tanks and Temples (6-view) | LPIPS: 0.1173 -> 0.1283 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 on most shared results.
- SparseCraft — SparseCraft: Few-Shot Neural Reconstruction through Stereopsis Guided Geometric Linearization (ECCV 2024) · better on 2 of 2 shared results
- UFORecon — UFORecon: Generalizable Sparse-View Surface Reconstruction from Arbitrary and Unfavorable Sets (CVPR 2024) · better on 1 of 1 shared result
- SPARF — SPARF: Neural Radiance Fields From Sparse and Noisy Poses (CVPR 2023) · 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
- Sparse2DGS: Geometry-Prioritized Gaussian Splatting for Surface Reconstruction from Sparse Views (CVPR 2025 · similarity 0.90)
- pixelSplat: 3D Gaussian Splats from Image Pairs for Scalable Generalizable 3D Reconstruction (CVPR 2024 · similarity 0.90)
- FatesGS: Fast and Accurate Sparse-View Surface Reconstruction Using Gaussian Splatting with Depth-Feature Consistency (AAAI 2025 · similarity 0.90)
- 3D Convex Splatting: Radiance Field Rendering with 3D Smooth Convexes (CVPR 2025 · similarity 0.90)
- Learning Efficient Fuse-and-Refine for Feed-Forward 3D Gaussian Splatting (NeurIPS 2025 · similarity 0.89)
- Generative Sparse-View Gaussian Splatting (CVPR 2025 · similarity 0.88)
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.
