AI Paper Atlas

Research map / 3D vision / Human pose estimation

Two-View Geometry Scoring Without Correspondences

Axel Barroso-Laguna, Eric Brachmann, Victor Adrian Prisacariu, Gabriel J. Brostow, Daniyar Turmukhambetov

CVPR 2023 · 3D vision › Human pose estimation · approach: Learned camera pose estimation · 9 citations (OpenAlex)

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 changedWhereMeasuredEffect
Score pose hypotheses with FSNet instead of MAGSAC++Fundamental matrix estimation · ScanNet test, 0-100 correspondencesmAA translation: 0.12 -> 0.19Improved
Score pose hypotheses with FSNet instead of MAGSAC++Essential matrix estimation · ScanNet test, 0-100 correspondencesmAA max(R,t): 0.14 -> 0.18Improved
Combine MAGSAC++ candidate selection with FSNet rerankingFundamental matrix estimation · ScanNet test, AllmAA max(R,t): 0.14 -> 0.19Improved
Combine MAGSAC++ candidate selection with FSNet rerankingEssential matrix estimation · PhotoTourism test, AllmAA translation: 0.65 -> 0.67Improved
Train using both fundamental- and essential-matrix hypotheses (ablation)FSNet · ScanNet test, fundamental estimation, AllmAA translation: 0.21 -> 0.20Hurt
Train using soft-clamped pose-error regression rather than binary cross-entropy classification (ablation)FSNet · ScanNet test, fundamental estimationmAA translation, 0-100: 0.18 -> 0.19Improved
Add soft clamping to the L1 pose-error loss (ablation)FSNet · ScanNet test, fundamental estimationmAA translation, 0-100: 0.16 -> 0.19Improved
Supervise with pose error rather than Sampson distance (ablation)FSNet · ScanNet test, fundamental estimationmAA rotation, 0-100: 0.23 -> 0.29Improved
Use a correspondence-count filter to select between FSNet and MAGSAC++ (ablation)Fundamental matrix estimation · ScanNet test, AllmAA max(R,t): 0.15 -> 0.16Improved
Loosen SuperGlue filtering to provide more correspondences (ablation)MAGSAC++ · ScanNet test, fundamental estimationmAA max(R,t): 0.14 -> 0.14No clear effect

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