AI Paper Atlas

Research map / 3D vision / Human pose estimation

Solving the Blind Perspective-n-Point Problem End-To-End With Robust Differentiable Geometric Optimization

Dylan Campbell, Liu Liu, Stephen Gould

ECCV 2020 · 3D vision › Human pose estimation · approach: Learned camera pose estimation

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
add pose loss to correspondence loss (ablation)BPnPNet (network output) · ModelNet40 testmedian rotation error (deg): 11.3 -> 9.66Improved
add pose loss to correspondence loss (ablation)BPnPNet (network output) · ModelNet40 testmedian translation error: 0.52 -> 0.08Improved
add pose loss to correspondence loss (ablation)BPnPNet (network output) · MegaDepth testmedian rotation error (deg): 4.47 -> 3.31Improved
add pose loss to correspondence loss (ablation)BPnPNet (network output) · MegaDepth testmedian translation error: 1.05 -> 0.46Improved
use in-network RANSAC pose estimate instead of network output (ablation)BPnPNet with pose loss · ModelNet40 testmedian rotation error (deg): 9.66 -> 8.09Improved
use in-network RANSAC pose estimate instead of network output (ablation)BPnPNet with pose loss · MegaDepth testmedian rotation error (deg): 3.31 -> 1.00Improved
add pose loss when RANSAC estimate is used (ablation)BPnPNet RANSAC output · ModelNet40 testmedian rotation error (deg): 11.7 -> 8.09Improved
add pose loss when RANSAC estimate is used (ablation)BPnPNet RANSAC output · MegaDepth testmedian rotation error (deg): 1.55 -> 1.00Improved
add pose loss to correspondence loss (ablation)BPnPNet (network output) · ModelNet40 testrecall at 15 deg rotation and 0.5 translation: 72%, an improvement of 25%Improved
add pose loss to correspondence loss (ablation)BPnPNet (network output) · MegaDepth testrecall at 10 deg rotation and 1 translation: 73%, 25% betterImproved
use in-network RANSAC pose estimate instead of network output (ablation)BPnPNet with pose loss · MegaDepth testrecall at 10 deg rotation and 1 translation: 73% -> 82%Improved
use end-to-end trained BPnPNet instead of SoftPOSITSoftPOSIT · MegaDepth testmedian rotation error (deg): 21.4 -> 1.00Improved

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