AI Paper Atlas

Research map / 3D vision / Human pose estimation

Eigendecomposition-free Training of Deep Networks with Zero Eigenvalue-based Losses

Zheng Dang, Kwang Moo Yi, Yinlin Hu, Fei Wang, Pascal Fua, Mathieu Salzmann

ECCV 2018 · 3D vision › Human pose estimation · approach: Learned camera pose estimation · 47 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
replace explicit eigendecomposition-based loss (Essential_Only) with the eigendecomposition-free loss (ablation)keypoint matching network · SUN3D and Strecha keypoint matchingmAPImproved
add classification-loss pre-training before the essential loss (ablation)eigendecomposition-free loss · SUN3D and Strecha keypoint matchingmAPNo clear effect
use eigendecomposition-free loss instead of SVD/Eigh gradients with vanilla gradient descent (ablation)plane fitting with one outlier · Plane fitting toy exampleiterations to convergeImproved
use eigendecomposition-free loss instead of SVD/Eigh gradients with Adam (ablation)plane fitting with multiple outliers · Plane fitting toy example (multiple outliers)inlier/outlier recoveryImproved
use eigendecomposition-free loss instead of explicit eigendecomposition loss (ablation)PnP weight network · Synthetic PnPtraining convergenceImproved
use eigendecomposition-free loss instead of REPPnP as outliers increasePnP weight network · Synthetic PnP (200 matches, outliers 10 to 150)rotation error: REPPnP degrades when there are more than 40 outliers; ours near zero error up to 130 outliersImproved
use eigendecomposition-free PnP network instead of EPnP+RANSACPnP weight network · PnP real images (Reichstag)rotation error: 3.1161 -> 0.1319Improved
use eigendecomposition-free PnP network instead of OPnPPnP weight network · PnP real images (Florence)rotation error: 1.5771 -> 0.0416Improved
use eigendecomposition-free PnP network instead of EPnP+RANSACPnP weight network · PnP real images (Notre-dame)translation error (normalized): 1.0605 -> 0.0046Improved
use eigendecomposition-free PnP network instead of REPPnPPnP weight network · PnP real images (Prague)translation error (normalized): 0.6156 -> 0.0069Improved

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