Research map / Optimization / Distributed and compressed
Distributed Very Large Scale Bundle Adjustment by Global Camera Consensus
Runze Zhang, Siyu Zhu, Tian Fang, Long Quan
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 over-relaxation (ablation) | camera consensus · Buildings | reprojection error (px): 1.35 -> 1.23 | Improved |
| add over-relaxation (ablation) | camera consensus · Town | reprojection error (px): 1.87 -> 1.76 | Improved |
| add over-relaxation (ablation) | camera consensus · Final 961 | iterations to converge: 30 -> 22 | Improved |
| add self-adaption of penalty (ablation) | camera consensus · Buildings | iterations to converge: 100 -> 64 | Improved |
| add self-adaption of penalty (ablation) | camera consensus · Roman Forum | reprojection error (px): 0.590 -> 0.536 | Improved |
| add self-adaption of penalty (ablation) | camera consensus · Venice | iterations to converge: 12 -> 16 | Hurt |
| add self-adaption of penalty (ablation) | camera consensus · Street | reprojection error (px): 1.05 -> 1.04 | No clear effect |
| camera consensus instead of point consensus | distributed BA · Town | data broadcast per iteration (MB): 342 -> 17.3 | Improved |
| camera consensus instead of point consensus | distributed BA · Final 13682 | data broadcast per iteration (MB): 106 -> 25.8 | Improved |
| camera consensus instead of point consensus | distributed BA · City | reprojection error (px): 1.41 -> 1.36 | Improved |
| camera consensus instead of point consensus | distributed BA · Venice | iterations to converge: 20 -> 16 | Improved |
| normalized-cut splitting instead of KD-tree splitting (ablation) | camera consensus · Final 13682 | cameras broadcast per iteration: 511k -> 282k | Improved |
Closest papers by meaning
- A Consensus-Based Framework for Distributed Bundle Adjustment (CVPR 2016 · similarity 0.92)
- Stochastic Bundle Adjustment for Efficient and Scalable 3D Reconstruction (ECCV 2020 · similarity 0.89)
- Very Large-Scale Global SfM by Distributed Motion Averaging (CVPR 2018 · similarity 0.88)
- Distributed Bundle Adjustment with Block-Based Sparse Matrix Compression for Super Large Scale Datasets (ICCV 2023 · similarity 0.88)
- Matrix-Free Shared Intrinsics Bundle Adjustment (CVPR 2025 · similarity 0.86)
- Stable and Scalable Bundle Adjustment of Holistic 3D Structures (ECCV 2026 · 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.
