AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Improving Face Recognition by Clustering Unlabeled Faces in the Wild

Aruni RoyChowdhury, Xiang Yu, Kihyuk Sohn, Erik Learned-Miller, Manmohan Chandraker

ECCV 2020 · Label-efficient and robust learning › Recognition and long-tail · approach: Fair and quality-aware face recognition · 17 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
add GCN-clustered pseudo-labeled unlabeled facesCosFace face recognition retrained with pseudo-labels · CFP-fpaccuracy: 92.37 -> 95.51Improved
add GCN-clustered pseudo-labeled unlabeled facesCosFace face recognition retrained with pseudo-labels · IJB-A verificationTAR@FAR=1e-3: 80.23 -> 87.60Improved
add GCN-clustered pseudo-labeled unlabeled facesCosFace face recognition retrained with pseudo-labels · IJB-A identificationRank-1: 92.66 -> 94.11Improved
add uncertainty-weighted loss with gamma=1 on pseudo-labels (ablation)CosFace face recognition retrained with pseudo-labels · IJB-A verificationTAR@FAR=1e-3: 87.60 -> 87.97Improved
vary gamma in uncertainty-weighted loss (1/3 to 3) (ablation)CosFace face recognition retrained with pseudo-labels · IJB-A, CFP-fpseveral: 87.93 to 88.43 TAR@FAR=1e-3 for gamma 1/3 to 3No clear effect
use gamma=1/2 instead of other gamma values (ablation)CosFace face recognition retrained with pseudo-labels · IJB-A verificationTAR@FAR=1e-3: 84.13 vs. 87.97 (gamma=1)Hurt
use Weibull 95% instead of naive Otsu threshold for overlapping identity separation (ablation)Max-logit OOD detection · Overlapping identity separationfalse positive rate: 6.2% -> 2.33%Improved
use Weibull instead of Gaussian model for overlap separation (ablation)Max-logit OOD detection · Overlapping identity separationfalse positive / negative rate: 2.01% vs. 2.33% FP, 0.51% vs. 0.50% FNNo clear effect
use classification-margin instead of softmax entropy as uncertainty metric (ablation)Noise model for clusters · Clustering error detectionaverage precision: 88.29% -> 95.16%Improved

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