Research map / 3D vision / NeRF and radiance fields
SG-NeRF: Neural Surface Reconstruction with Scene Graph Optimization
Yiyang Chen, Siyan Dong, Xulong Wang, Lulu Cai, Youyi Zheng, Yanchao Yang
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 confidence-score-based sampling of images (ablation) | SG-NeRF · Proposed dataset (3 scenes) | Chamfer distance: 0.67 -> 0.47 | Improved |
| add IoU loss on matched keypoints (ablation) | SG-NeRF · Proposed dataset (3 scenes) | Chamfer distance: 0.72 -> 0.47 | Improved |
| add angular-threshold graph pruning (ablation) | SG-NeRF · Proposed dataset (3 scenes) | Chamfer distance: 0.60 -> 0.47 | Improved |
| add coarse-to-fine Gaussian smoothing (ablation) | SG-NeRF · Proposed dataset (3 scenes) | Chamfer distance: 0.58 -> 0.47 | Improved |
| use SG-NeRF instead of plain NeuS under outlier poses | NeuS · Proposed outlier-pose dataset | Chamfer distance: 1.18 -> 0.46 | Improved |
| add local pose optimizers (BARF, SCNeRF, GARF, L2G, Joint-TensoRF) to NeuS | NeuS · Proposed outlier-pose dataset | Chamfer distance: 1.18 -> 1.05 (BARF*), 1.64 (GARF*) | Mixed |
| add pose optimization to Neuralangelo instead of NeuS | Neuralangelo · Proposed outlier-pose dataset | Chamfer distance: BARF# 2.34 vs BARF* 1.05 | Hurt |
| run bundle adjustment on the pruned scene graph | NeuS · Proposed outlier-pose dataset | pose update | No clear effect |
| increase lambda from 0.2 to 1.0, then to 2.0 and 3.0 (ablation) | SG-NeRF · Proposed dataset | Chamfer distance: 0.5058 -> 0.4646 -> 0.4885 | Mixed |
Later papers that compared with this method
1 later accepted paper reports SG-NeRF as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- ManifoldNeuS: Manifold-aware View Optimizability for Pose-Free Neural Surface Reconstruction (CVPR 2026 · better on 1 of 1 shared result)
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- NeuS — NeuS: Learning Neural Implicit Surfaces by Volume Rendering for Multi-view Reconstruction (NeurIPS 2021) · better on 3 of 3 shared results
- BARF* — BARF: Bundle-Adjusting Neural Radiance Fields (ICCV 2021) · better on 3 of 3 shared results
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
- Neural Radiance Fields From Sparse RGB-D Images for High-Quality View Synthesis (TPAMI 2022 · similarity 0.89)
- Aug-NeRF: Training Stronger Neural Radiance Fields With Triple-Level Physically-Grounded Augmentations (CVPR 2022 · similarity 0.89)
- A Probability-guided Sampler for Neural Implicit Surface Rendering (ECCV 2024 · similarity 0.89)
- NeRFPrior: Learning Neural Radiance Field as a Prior for Indoor Scene Reconstruction (CVPR 2025 · similarity 0.89)
- ProvNeRF: Modeling per Point Provenance in NeRFs as a Stochastic Field (NeurIPS 2024 · similarity 0.88)
- ICON: Incremental CONfidence for Joint Pose and Radiance Field Optimization (CVPR 2024 · 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.
