Research map / 3D vision / Depth and stereo
Learning Non-Volumetric Depth Fusion Using Successive Reprojections
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 |
|---|---|---|---|
| Add successive reprojection of twelve mixed close and far neighbors to COLMAP depth estimates | DeFuSR on COLMAP · DTU full point cloud | comp. (%): 72 -> 84 | Helped |
| Add successive reprojection of twelve mixed close and far neighbors to COLMAP depth estimates | DeFuSR on COLMAP · DTU per-view | acc. (%): 66 -> 89 | Helped |
| Use zero neighbors rather than neighboring-view information (ablation) | DeFuSR on COLMAP · DTU full point cloud | comp. (%): 66 -> 84 | Hurt |
| Use four neighbors instead of twelve (ablation) | DeFuSR on COLMAP · DTU full point cloud | comp. (%): 64 -> 84 | Hurt |
| Increase the neighbor set from four to twelve (ablation) | DeFuSR on COLMAP · DTU full point cloud | comp. (%): 64 -> 84 | Helped |
| Refine MVSNet depth estimates | DeFuSR on MVSNet · DTU per-view | acc. (%): 76 -> 92 | Helped |
| Refine MVSNet depth estimates | DeFuSR on MVSNet · DTU full point cloud | acc. (%): 88 -> 86 | Hurt |
| Refine MVSNet depth estimates | DeFuSR on MVSNet · DTU full point cloud | comp. (%): 66 -> 65 | Hurt |
| Use mixed close and far neighbors instead of only nearest neighbors (ablation) | DeFuSR on COLMAP · DTU full point cloud | comp. (%): 72 -> 84 | Helped |
| Use mixed close and far neighbors instead of only furthest neighbors (ablation) | DeFuSR on COLMAP · DTU full point cloud | comp. (%): 76 -> 84 | Helped |
Closest papers by meaning
- Differentiable Diffusion for Dense Depth Estimation From Multi-View Images (CVPR 2021 · similarity 0.89)
- Deep Stereo Using Adaptive Thin Volume Representation With Uncertainty Awareness (CVPR 2020 · similarity 0.89)
- MVD-Fusion: Single-view 3D via Depth-consistent Multi-view Generation (CVPR 2024 · similarity 0.88)
- Zero-Shot Novel View and Depth Synthesis with Multi-View Geometric Diffusion (CVPR 2025 · similarity 0.88)
- RRT-MVS: Recurrent Regularization Transformer for Multi-View Stereo (AAAI 2025 · similarity 0.88)
- Depth-Guided Robust and Fast Point Cloud Fusion NeRF for Sparse Input Views (AAAI 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.