Research map / 3D vision / Depth and stereo
Multi-View Depth Estimation by Fusing Single-View Depth Probability With Multi-View Geometry
Check 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 |
|---|---|---|---|
| Fuse single-view depth probability with multi-view geometry using probabilistic sampling and consistency weighting | MaGNet · KITTI Eigen split | abs rel: 0.061 -> 0.054 | Helped |
| Fuse single-view depth probability with multi-view geometry | MaGNet · ScanNet | NLL: 2.235 -> 0.145 | Helped |
| Fuse single-view depth probability with multi-view geometry | MaGNet · 7-Scenes | NLL: 2.794 -> 1.561 | Helped |
| Fuse single-view depth probability with multi-view geometry | MaGNet · KITTI | NLL: 2.644 -> 1.805 | Helped |
| Increase the number of candidates from 5 to 11 in a single iteration (ablation) | MaGNet · ScanNet ablation test set | rmse: 0.217 -> 0.216 | No clear effect |
| Repeat multi-view matching for 3 rather than 1 iteration, with 5 candidates per iteration (ablation) | MaGNet · ScanNet ablation test set | rmse: 0.217 -> 0.207 | Helped |
| Use 4 rather than 3 matching iterations (ablation) | MaGNet · ScanNet ablation test set | rmse: 0.207 -> 0.206 | Helped |
| Replace UCS-Net's initial uniform sampling with probabilistic sampling (ablation) | UCS-Net · ScanNet ablation test set | rmse: 0.229 -> 0.214 | Helped |
| Use probabilistic sampling plus consistency weighting rather than uniform sampling without the proposed fusion (ablation) | Multi-view matching · ScanNet ablation test set | RMSE | Helped |
| Add consistency weighting to uniformly sampled candidates (ablation) | Multi-view depth estimation · ScanNet ablation test set | accuracy | Helped |
Closest papers by meaning
- MVSAnywhere: Zero-Shot Multi-View Stereo (CVPR 2025 · similarity 0.87)
- Adaptive Fusion of Single-View and Multi-View Depth for Autonomous Driving (CVPR 2024 · similarity 0.86)
- Depth2Pose: A Pose-Based Benchmark for Monocular Depth Estimation without Ground-Truth Depth (NeurIPS 2026 · similarity 0.86)
- Multi-view Reconstruction via SfM-guided Monocular Depth Estimation (CVPR 2025 · similarity 0.86)
- Depth Pro: Sharp Monocular Metric Depth in Less Than a Second (ICLR 2025 · similarity 0.86)
- DME: Unveiling the Bias for Better Generalized Monocular Depth Estimation (AAAI 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.