Research map / Learning theory and algorithms / Matrix and tensor methods
Shonan Rotation Averaging: Global Optimality by Surfing SO(p)(n)
Frank Dellaert, David M. Rosen, Jing Wu, Robert Mahony, Luca Carlone
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 Shonan Averaging instead of Levenberg-Marquardt | rotation averaging · Synthetic n=20 sigma=0.5 | success rate: 20% -> 100% | Improved |
| use Shonan Averaging instead of block coordinate descent | rotation averaging · Synthetic n=200 sigma=0.2 | success rate: 0% -> 100% | Improved |
| use Shonan Averaging instead of block coordinate descent | rotation averaging · Synthetic n=100 sigma=0.5 | time (s): 32.982 -> 0.373 | Improved |
| use Shonan Averaging instead of Levenberg-Marquardt | rotation averaging · YFCC reichstag | success rate: 80% -> 100% | Improved |
| use Levenberg-Marquardt instead of Shonan Averaging for speed | rotation averaging · YFCC pantheon_interior | time (s): 0.971 -> 0.313 | Improved |
| start Shonan Averaging from p=3 instead of p=5 (ablation) | Shonan Averaging · YFCC statue_of_liberty_2 | average time (s): 0.063 -> 0.034 | Improved |
| add Karcher mean gauge prior (ablation) | Shonan Averaging · YFCC statue_of_liberty_1 | average time (s): 0.313 -> 0.108 | Improved |
| add Karcher mean gauge prior (ablation) | Shonan Averaging · YFCC taj_mahal_entrance | average time (s): 0.117 -> 0.081 | Improved |
| raise relaxation rank from p=3 to p=4 (ablation) | single-level Shonan · YFCC natural_history_museum_london | success rate: 60% -> 100% | Improved |
| raise relaxation rank from p=3 to p=5 (ablation) | single-level Shonan · YFCC statue_of_liberty_1 | success rate: 0% -> 40% | Improved |
Closest papers by meaning
- Rotation Coordinate Descent for Fast Globally Optimal Rotation Averaging (CVPR 2021 · similarity 0.91)
- Rotation Averaging with the Chordal Distance: Global Minimizers and Strong Duality (TPAMI 2019 · similarity 0.90)
- Certifiably Optimal Anisotropic Rotation Averaging (ICCV 2025 · similarity 0.89)
- Rotation Averaging in a Split Second: A Primal-Dual Method and a Closed-Form for Cycle Graphs (ICCV 2021 · similarity 0.89)
- Hybrid Rotation Averaging: A Fast and Robust Rotation Averaging Approach (CVPR 2021 · similarity 0.88)
- Revisiting Rotation Averaging: Uncertainties and Robust Losses (CVPR 2023 · similarity 0.87)
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.
