Research map / 3D vision / Gaussian splatting
Learning-Based Multi-View Stereo: A Survey
Fangjinhua Wang, Qingtian Zhu, Di Chang, Quankai Gao, Junlin Han, Tong Zhang, Richard Hartley, Marc Pollefeys
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 offline depth-map MVS instead of online depth-map MVS | depth-map MVS · Tanks and Temples (Ignatius) | F-score: 0.01 -> 0.81 | Improved |
| use depth-map representation instead of voxel representation | MVS reconstruction · Tanks and Temples (Truck) | F-score: 0.15 -> 0.70 | Improved |
| scale NeRF-based method from small to large scene (Ignatius to Courthouse) | Neuralangelo · Tanks and Temples | F-score: 0.89 -> 0.28 | Hurt |
| scale 3DGS-based method from small to large scene (Ignatius to Courthouse) | PGSR · Tanks and Temples | F-score: 0.80 -> 0.21 | Hurt |
| use feed-forward point-based reconstruction instead of offline depth-map MVS | MapAnything · Tanks and Temples (mean) | F-score: 0.60 -> 0.19 | Hurt |
| add geometry regularization to 3DGS (PGSR vs SuGaR) | 3DGS-based reconstruction · Tanks and Temples (mean) | F-score: 0.19 -> 0.50 | Improved |
| use NeRF with multi-resolution hash grid (Neuralangelo) over NeuS | NeuS · Tanks and Temples (mean) | F-score: 0.38 -> 0.50 | Improved |
| add transformer/ViT features and refinement (MVSFormer++ over CasMVSNet) | CasMVSNet · Tanks and Temples (mean) | F-score: 0.49 -> 0.59 | Improved |
Closest papers by meaning
- MVS2D: Efficient Multi-View Stereo via Attention-Driven 2D Convolutions (CVPR 2022 · similarity 0.93)
- Fast-MVSNet: Sparse-to-Dense Multi-View Stereo With Learned Propagation and Gauss-Newton Refinement (CVPR 2020 · similarity 0.92)
- Just a Few Points Are All You Need for Multi-View Stereo: A Novel Semi-Supervised Learning Method for Multi-View Stereo (ICCV 2021 · similarity 0.92)
- Zero-Shot Novel View and Depth Synthesis with Multi-View Geometric Diffusion (CVPR 2025 · similarity 0.90)
- Multi-view Reconstruction via SfM-guided Monocular Depth Estimation (CVPR 2025 · similarity 0.90)
- MVSAnywhere: Zero-Shot Multi-View Stereo (CVPR 2025 · similarity 0.89)
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.
