AI Paper Atlas

Research map / Learning theory and algorithms / Prediction and decision losses

Repeatability Is Not Enough: Learning Affine Regions via Discriminability

Dmytro Mishkin, Filip Radenovic, Jiri Matas

ECCV 2018 · Learning theory and algorithms › Prediction and decision losses · approach: Consistency of surrogate losses · 246 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 hardest-negative gradient (HardNeg) with HardNegC (ablation)AffNet, HardNet descriptor · HSequences hardest pairscorrect matches: n/c -> 388Improved
use HardNegC instead of PosDist loss for affine shape (ablation)AffNet, SIFT descriptor · HSequences hardest pairscorrect matches: n/c -> 386Improved
learn affine shape with AffNet instead of BaumbergHessian + HardNet · HSequences hardest pairscorrect matches: 298 -> 388Improved
learn orientation instead of dominant gradient orientationHessian + HardNet · HSequences hardest pairscorrect matches: 339 -> 387Improved
predict residual shape A'' with zero-bias parametrization instead of full matrix A (ablation)AffNet · HPatchesSeqcorrect matches: 360 -> 388Improved
learn orientation separately after shape instead of jointly (ablation)AffNet with A' and R(alpha) · HPatchesSeqcorrect matches: 250 -> 327Improved
learn orientation instead of dominant gradient (ablation)AffNet A'' parameterization · HPatchesSeqcorrect matches: 349 -> 388Improved
use AffNet instead of BaumbergDoG detector · HPatches viewpointrepeatability: 0.51 -> 0.52Improved
use AffNet instead of BaumbergHessian detector · HPatches viewpointrepeatability: 0.52 -> 0.56Improved
replace Baumberg with AffNetHardNet++ BoW retrieval · Oxford5k BoWmAP: 60.8 -> 68.3Improved
replace Baumberg with AffNetTFeat-M* BoW retrieval · Oxford5k BoWmAP: 46.7 -> 45.5Mixed
keep all AffNet features instead of restricting to Baumberg count (ablation)HesAffNet-HardNet++ · Oxford5k BoWmAP: 64.3 -> 68.3Improved

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