atlas

Research map / 3D vision / Depth and stereo

Hybrid Rotation Averaging: A Fast and Robust Rotation Averaging Approach

CVPR 2021 · 3D vision › Depth and stereo · approach: Feed-forward 3D reconstruction · 32 citations (OpenAlex)

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 changedWhereMeasuredEffect
Use RA-SfM with hybrid rotation averaging and averaged-rotation regularization instead of COLMAPIncremental SfM · TrafalgarMRE: 0.81 -> 0.72Helped
Use RA-SfM instead of COLMAPIncremental SfM · Union SquareMRE: 0.66 -> 0.52Helped
Use RA-SfM instead of COLMAPIncremental SfM · PiccadillyMRE: 0.73 -> 0.80Hurt
Use RA-SfM instead of COLMAPIncremental SfM · Roman ForumMRE: 0.71 -> 0.77Hurt
Use RA-SfM instead of COLMAPIncremental SfM · Online datasetsMREMixed
Add fast view graph filtering before global rotation averaging (ablation)Global rotation averaging · Synthetic outlier experimentsmean rotation errorHelped
Add local IRLS refinement after global rotation averaging (ablation)Global rotation averaging · Synthetic outlier experimentsmean rotation errorHelped
Use hybrid RA instead of SE-SyncRotation averaging · Synthetic, n=50000, σ=0.2time (s): 956.821 -> 0.515Helped
Add averaged rotations as regularizers in bundle adjustmentIncremental SfM · Campusloop closureHelped

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy