AI Paper Atlas

Research map / Graph learning / Graph algorithms and clustering

Adaptive Annealing for Robust Averaging

Sidhartha Chitturi, Venu Madhav Govindu

ECCV 2024 · Graph learning › Graph algorithms and clustering · approach: Graph comparison and random features · 1 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 fixed-factor annealing with adaptive Hessian-guided annealing (GNCp)GNC for vector averaging · Vector averaging synthetic, outlier fraction 0.5mean error: 13.99 -> 1.37 vs GNC (delta down)Improved
use adaptive annealing instead of large fixed delta (delta=5)GNC for vector averaging · Vector averaging, outlier fraction 0.5mean error: 86.26 -> 1.37Improved
adaptive annealing vs small fixed deltaGNC for vector averaging · Vector averaging, outlier fraction 0.5time (s): 15.37 -> 4.9Improved
large fixed delta annealing vs adaptive annealingGNC for vector averaging · Vector averaging, outlier fraction 0.5time (s): 3.56 vs 4.9 (fast but poor accuracy)Hurt
adaptive annealing at low outlier fractionGNC for vector averaging · Vector averaging, outlier fraction 0.1mean error: 0.41 vs 0.41No clear effect
adaptive annealing on real indoor dataAugmented ICL-NUIM · Augmented ICL-NUIMmean error: 1.93 vs 1.93 (lr1)No clear effect
adaptive annealing on outlier filteringGNC for vector averaging · Vector averaging, outlier fraction 0.5outliers remaining (%): 55.2 -> 0Improved
GNCp-LUD with Geman-McClure vs original LUD on real 1DSfM dataTranslation averaging · 1DSfM datasetsmean/median error: superior in 9 (mean) and 11 (median) of 15 datasetsMixed
GNCp-LUD on TrafalgarTranslation averaging · 1DSfM Trafalgarmean error: 8.4 -> 10.4 vs LUDHurt

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