AI Paper Atlas

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

Robustness Disparities in Face Detection

Samuel Dooley, George Z Wei, Tom Goldstein, John P Dickerson

NeurIPS 2022 · Label-efficient and robust learning › Recognition and long-tail · approach: Fair and quality-aware face recognition · 2 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
compare commercial against academic modelscommercial face detectors · all four datasetsaverage precision under corruptionImproved
use MogFace instead of commercial modelsMogFace · UTKFaceaverage precision under corruptionImproved
use MogFace instead of TinaFace and Yolo5FaceMogFace · all datasets except CCDaverage precision under corruptionImproved
apply zoom blur corruptionall six models · CCD and MIAPaverage precisionHurt
apply elastic transform corruptionfour of six models · UTKFaceaverage precisionHurt
apply noise corruptionsTinaFace and Yolo5Face · UTKFaceaverage precisionHurt
perceived feminine vs masculine presentationsix face detectors · 24 dataset-model combinationsnoise-induced AP change: 19 of 24 biased against masculine; 4 against feminine; 1 noneMixed
darker vs lighter skin typeall models · CCDnoise robustness APHurt
dim vs bright lightingall models except YOLO5Face · CCDnoise robustness APHurt
dim lighting on skin type gapsix face detectors · CCDskin-type AP gapImproved
older vs middle-aged subjectssix face detectors · Adience, MIAP, UTKFacenoise robustness APHurt
commercial vs academic models in disparity magnitudecommercial face detectors · all datasetsdemographic disparityNo clear effect

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