AI Paper Atlas

Research map / Image generation / Editing and style transfer

Goldilocks Test Sets for Face Verification

Haiyu Wu, Sicong Tian, Aman Bhatta, Jacob Gutierrez, Grace Bezold, Genesis Argueta, Karl Ricanek, Michael C. King, Kevin Bowyer

CVPR 2026 · Image generation › Editing and style transfer · approach: Identity-preserving face manipulation · 3 citations (OpenAlex)

Read the paper

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
use random MORPH5 pairs instead of attribute-targeted pairs (ablation)Hadrian/Eclipse · Random MORPH5 test setaccuracy: 99.82 vs 91.72 (Hadrian) and 81.73 (Eclipse), ArcFace MS1MV2Hurt
assign genuine pairs of same identity to different folds (overlap) instead of disjoint (ablation)Hadrian · Hadrianaccuracy: 91.72 -> 91.72 (ArcFace), 91.48 -> 91.45 (CurricularFace)No clear effect
identity overlap across folds instead of disjoint (ablation)Eclipse · Eclipseaccuracy: 81.73 -> 81.68 (ArcFace), 82.67 -> 82.60 (MagFace)No clear effect
identity overlap across folds instead of disjoint (ablation)ND-Twins · ND-Twinsaccuracy: 68.55 -> 68.70 (ArcFace), 65.73 -> 65.82 (UniFace)No clear effect
compare African-American male vs Caucasian male subsetsArcFace · Hadrianaccuracy: 97.17 (AAM) vs 86.27 (CM)Mixed
compare Caucasian female vs African-American male subsetsArcFace · Eclipseaccuracy: 95.00 (AAM) vs 65.33 (CF)Mixed
train on larger/cleaner Glint360K instead of MS1MV2ArcFace · ND-Twinsaccuracy: 68.55 -> 84.67Improved
train on Glint360K instead of MS1MV2ArcFace · Hadrianaccuracy: 91.72 -> 95.68Improved
do not restrict image occurrence across pairs (ablation)FR models · ND-Twinsaccuracy: none better than 50% without constraint vs 71.56 average with itHurt
do not restrict image occurrence across pairs (ablation)FR models · Hadrian/Eclipseaccuracy: less than 55% without constraint vs 92.62 and 82.81 average with itHurt

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