Research map / 3D vision / Gaussian splatting
Generalizable Neural Reconstruction of High-Fidelity Surfaces via Sparse Volumetric Representations
Aoxiang Fan, Corentin Dumery, Nicolas Talabot, Ming Xu, Hieu Le, Pascal Fua
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 |
|---|---|---|---|
| use sparse 512^3 volumes instead of 128^3 (ablation) | SVRecon · DTU | Chamfer distance: 1.27 -> 1.00 | Improved |
| raise resolution from 256^3 to 512^3 (ablation) | SVRecon · DTU | Chamfer distance: 1.04 -> 1.00 | Improved |
| use 32 instead of 16 base feature channels (ablation) | SVRecon · DTU | Chamfer distance: 1.15 -> 1.00 | Improved |
| use more input views at test time (3 to 5) (ablation) | SVRecon · DTU | Chamfer distance: 1.00 -> 0.96 | Improved |
| lower occupancy threshold (0.5 to 0.1) to favor recall (ablation) | SVRecon · DTU | Chamfer distance: 1.03 -> 1.00 | Improved |
| larger dilation kernel (3 to 7) (ablation) | SVRecon · DTU | Chamfer distance: 1.00 -> 0.99 | No clear effect |
| sparse occupancy-driven volumes versus ReTR dense baseline | SVRecon · DTU | Normal consistency AUC@15: 16.3 -> 21.0 | Improved |
| compare against UFORecon on Chamfer distance | SVRecon · DTU | Chamfer distance: 1.00 vs 1.00 | No clear effect |
| generalize to unseen Co3D versus MVSFormer++ | SVRecon · Co3D | AUC@0.05: 0.60 -> 0.72 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 4 of 5 on most shared results.
- UFORecon — UFORecon: Generalizable Sparse-View Surface Reconstruction from Arbitrary and Unfavorable Sets (CVPR 2024) · better on 1 of 2 shared results
- SparseCraft — SparseCraft: Few-Shot Neural Reconstruction through Stereopsis Guided Geometric Linearization (ECCV 2024) · better on 2 of 2 shared results
- ReTR — ReTR: Modeling Rendering Via Transformer for Generalizable Neural Surface Reconstruction (NeurIPS 2023) · better on 2 of 2 shared results
- MVSFormer++ — MVSFormer++: Revealing the Devil in Transformer's Details for Multi-View Stereo (ICLR 2024) · better on 1 of 1 shared result
- VolRecon — VolRecon: Volume Rendering of Signed Ray Distance Functions for Generalizable Multi-View Reconstruction (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
- Surface-Centric Modeling for High-Fidelity Generalizable Neural Surface Reconstruction (ECCV 2024 · similarity 0.90)
- GeoSVR: Taming Sparse Voxels for Geometrically Accurate Surface Reconstruction (NeurIPS 2025 · similarity 0.88)
- UFORecon: Generalizable Sparse-View Surface Reconstruction from Arbitrary and Unfavorable Sets (CVPR 2024 · similarity 0.88)
- NeuSurf: On-Surface Priors for Neural Surface Reconstruction from Sparse Input Views (AAAI 2024 · similarity 0.87)
- SurfelSplat: Learning Efficient and Generalizable Gaussian Surfel Representations for Sparse-View Surface Reconstruction (NeurIPS 2025 · similarity 0.86)
- GVKF: Gaussian Voxel Kernel Functions for Highly Efficient Surface Reconstruction in Open Scenes (NeurIPS 2024 · similarity 0.86)
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.
