Research map / 3D vision / Depth and stereo
GenS: Generalizable Neural Surface Reconstruction from Multi-View Images
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Replace SparseNeuS with GenS | Sparse-input reconstruction on DTU · DTU sparse inputs | Chamfer Distance: 1.64 -> 1.34 | Helped |
| Fine-tune GenS | GenS on DTU · DTU sparse inputs | Chamfer Distance: 1.34 -> 1.03 | Helped |
| Add multi-scale feature-metric consistency (MFC) (ablation) | SparseNeuS baseline · DTU sparse inputs | Chamfer Distance: 1.90 -> 1.62 | Helped |
| Add generalized multi-scale volume (GMV) (ablation) | MFC-equipped SparseNeuS baseline · DTU sparse inputs | Chamfer Distance: 1.62 -> 1.42 | Helped |
| Add view contrast loss (VCL) (ablation) | MFC + GMV model · DTU sparse inputs | Chamfer Distance: 1.42 -> 1.34 | Helped |
| Add patch similarity (ablation) | MFC · DTU | Chamfer Distance: 1.86 -> 1.76 | Helped |
| Add multi-scale features to patch similarity (ablation) | MFC · DTU | Chamfer Distance: 1.76 -> 1.73 | Helped |
| Add self-enhancement to multi-scale patch similarity (ablation) | MFC · DTU | Chamfer Distance: 1.73 -> 1.62 | Helped |
| Fine-tune GenS | GenS on DTU dense inputs · DTU dense inputs | Chamfer Distance: 0.80 -> 0.62 | Helped |
Closest papers by meaning
- UFORecon: Generalizable Sparse-View Surface Reconstruction from Arbitrary and Unfavorable Sets (CVPR 2024 · similarity 0.88)
- 3D Reconstruction with Generalizable Neural Fields using Scene Priors (ICLR 2024 · similarity 0.88)
- SC-NeuS: Consistent Neural Surface Reconstruction from Sparse and Noisy Views (AAAI 2024 · similarity 0.88)
- UniSDF: Unifying Neural Representations for High-Fidelity 3D Reconstruction of Complex Scenes with Reflections (NeurIPS 2024 · similarity 0.88)
- PMNI: Pose-free Multi-view Normal Integration for Reflective and Textureless Surface Reconstruction (CVPR 2025 · similarity 0.87)
- SuperNormal: Neural Surface Reconstruction via Multi-View Normal Integration (CVPR 2024 · similarity 0.87)
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.