AI Paper Atlas

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

Meta Balanced Network for Fair Face Recognition

Mei Wang, Yaobin Zhang, Weihong Deng

TPAMI 2021 · Label-efficient and robust learning › Recognition and long-tail · approach: Fair and quality-aware face recognition · 72 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
use MBN(arc) instead of ArcfaceMBN · BUPT-Globalface-4 / IDS-4SER: 2.33 -> 1.59Improved
use MBN(cos) instead of CosfaceMBN · BUPT-Globalface-8 / IDS-8SER: 2.28 -> 1.55Improved
use MBN(arc) instead of ArcfaceMBN · BUPT-Balancedface-4 / IDS-4SER: 1.65 -> 1.37Improved
learn adaptive margins per skin-tone group (MBN(arc)) instead of uniform marginMBN · BUPT-Globalface-4 / IDS-4average accuracy: 95.37 -> 95.93Improved
MBN(soft) vs Norm-Softmax on skewed training distribution 7:1:1:1MBN · simulated Globalface-4STD: 3.46 -> 2.19Improved
MBN(soft) vs Norm-Softmax on training distribution 4:2:2:2MBN · simulated Globalface-4STD: 2.64 -> 1.33Improved
train on skin-tone-balanced data instead of GlobalfaceArcface · IDS-8STD: 1.52 -> 1.16Improved
train on skin-tone-balanced Balancedface* instead of CASIA-WebFaceArcface · IDS-4 Tone-IVaccuracy: 84.93 -> 90.98Improved
use sample re-weighting instead of MBNArcface · BUPT-Globalface-4 / IDS-4SER: 1.79 -> 1.59Hurt
use adversarial attribute removal instead of MBNArcface · BUPT-Globalface-4 / IDS-4SER: 1.87 -> 1.59Hurt
use MBN instead of RL-RBN for margin learningMBN(arc) · BUPT-Globalface-4 / IDS-4SER: 1.76 -> 1.59Hurt
use MBN instead of Arcface on age cohorts of Tone-VIMBN · IDS-4 Tone-VI age cohortsSTD: 3.21 -> 0.94Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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