Research map / Graph learning / Graph algorithms and clustering
Bundle Adjustment on a Graph Processor
Joseph Ortiz, Mark Pupilli, Stefan Leutenegger, Andrew J. Davison
Check 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 GBP on the IPU instead of Ceres for bundle adjustment | Bundle adjustment · 10 TUM sequences | time to ARE < 1.5 pixels: 24x faster on average | Improved |
| Use GBP on the IPU instead of Ceres after each new keyframe | Incremental SLAM · First 30 keyframes of three TUM sequences | time to ARE < 1.5 pixels: 36x faster on average | Improved |
| Add a Huber loss to GBP (ablation) | GBP · fr1desk with incorrect data associations | converged reprojection error | Improved |
| Remove the Huber loss (ablation) | GBP · fr1desk with more than 3% bad associations | convergence | Hurt |
| Add a Huber loss (ablation) | Ceres · fr1desk with incorrect data associations | convergence | Mixed |
| Use GBP instead of Ceres for convergence from noisy keyframe initializations | Bundle adjustment · fr1desk and fr3teddy | successful convergence rate | No clear effect |
Closest papers by meaning
- Learning To Bundle-Adjust: A Graph Network Approach to Faster Optimization of Bundle Adjustment for Vehicular SLAM (ICCV 2021 · similarity 0.86)
- Belief Propagation Converges to Gaussian Distributions in Sparsely-Connected Factor Graphs (ICML 2026 · similarity 0.85)
- Parallel Rigidity Matters for Bundle Adjustment (CVPR 2026 · similarity 0.84)
- Graph Cuts Always Find a Global Optimum for Potts Models (With a Catch) (ICML 2021 · similarity 0.84)
- Parallelizable Riemannian Alternating Direction Method of Multipliers for Non-convex Pose Graph Optimization (AAAI 2026 · similarity 0.83)
- A machine learning approach that beats Rubik's cubes (NeurIPS 2025 · similarity 0.82)
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.
