Research map / 3D vision / Depth and stereo
Correspondence Reweighted Translation Averaging
Lalit Manam, Venu Madhav Govindu
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 correspondence reweighting (CReTA) on top of RLUD | RLUD · MNDsyn sigma=3 with outliers | mean translation error (m): 1.49 -> 0.19 | Improved |
| add correspondence reweighting (CReTA) on top of BATA | BATA · MNDsyn sigma=3 with outliers | mean translation error (m): 0.29 -> 0.25 | Improved |
| add correspondence reweighting (CReTA) on top of RLUD | RLUD · TOLsyn sigma=3 with outliers | mean translation error (m): 12.47 -> 7.89 | Improved |
| add CReTA refinement to input translation directions | RLUD · TOLsyn sigma=3 with outliers | mean relative direction error (deg): 3.60 -> 1.05 | Improved |
| add CReTA on top of BATA with own initialization | BATA · 1DSfM Madrid Metropolis | mean translation error (m): 6.9 -> 6.1 | Improved |
| add CReTA on top of RLUD with own initialization | RLUD · 1DSfM Notre Dame | mean translation error (m): 1.2 -> 2.1 | Hurt |
| add CReTA on top of BATA with weak 1DSfM-provided directions | BATA · 1DSfM Yorkminster | mean translation error (m): 9.4 -> 5.2 | Improved |
| add CReTA on top of BATA with weak 1DSfM-provided directions | BATA · 1DSfM Montreal Notre Dame | mean translation error (m): 1.3 -> 2.6 | Hurt |
| add CReTA to translation averaging before triangulation | LUD · 1DSfM Alamo triangulation | mean reprojection error (px): 6.6 -> 3.0 | Improved |
| add removal of low quality correspondences after first iteration (ablation) | CReTA-RLUD · 1DSfM Vienna Cathedral | mean translation error (m): 11.4 -> 7.3 | Improved |
| add removal of low quality correspondences after first iteration (ablation) | CReTA-RLUD · 1DSfM Union Square | mean translation error (m): 26.3 -> 15.9 | Improved |
| add removal of low quality correspondences after first iteration (ablation) | CReTA-RLUD · 1DSfM Ellis Island | mean translation error (m): 28.3 -> 9.5 | Improved |
Closest papers by meaning
- Sensitivity in Translation Averaging (NeurIPS 2023 · similarity 0.92)
- Baseline Desensitizing in Translation Averaging (CVPR 2018 · similarity 0.90)
- Revisiting Global Translation Estimation with Feature Tracks (CVPR 2024 · similarity 0.84)
- MultiDiff: Consistent Novel View Synthesis from a Single Image (CVPR 2024 · similarity 0.83)
- FRESCO: Spatial-Temporal Correspondence for Zero-Shot Video Translation (CVPR 2024 · similarity 0.83)
- SynthRGB-T: Language-Vision Guided Image Translation for Diversity Synthesis (CVPR 2026 · 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.
