AI Paper Atlas

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

A realistic approach to generate masked faces applied on two novel masked face recognition data sets

Tudor-Alexandru Mare, Georgian Duta, Iuliana Georgescu, Adrian Sandru, Bogdan Alexe, Marius Popescu, Radu Tudor Ionescu

NeurIPS 2021 · Label-efficient and robust learning › Recognition and long-tail · approach: Fair and quality-aware face recognition · 4 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
fine-tune on CelebA with synthetic masks instead of unmasked imagesFaceNet · CelebA-test+masksverification accuracy: 90.96 -> 93.58Improved
fine-tune on CelebA with synthetic masks instead of unmasked imagesVGG-face · CelebA-test+masksverification accuracy: 84.56 -> 91.51Improved
fine-tune on CelebA with synthetic masks instead of unmasked imagesArcFace · CelebA-test+masksverification accuracy: 91.78 -> 95.43Improved
fine-tune on CASIA-WebFace with synthetic masks instead of unmasked imagesFaceNet · CASIA-WebFace-test+masksverification accuracy: 84.21 -> 88.06Improved
fine-tune on CASIA-WebFace with synthetic masks instead of unmasked imagesVGG-face · CASIA-WebFace-test+masksverification accuracy: 79.65 -> 86.85Improved
fine-tune on CASIA-WebFace with synthetic masks instead of unmasked imagesArcFace · CASIA-WebFace-test+masksverification accuracy: 87.95 -> 91.47Improved
train on our masks instead of unmasked CelebAFaceNet · MFR2verification accuracy: 92.20 -> 96.22Improved
train on our masks instead of unmasked CelebAVGG-face · MFR2verification accuracy: 84.29 -> 91.26Improved
replace MaskTheFace masks with our SparkAR masks in training (ablation)VGG-face · MFR2verification accuracy: 88.35 -> 91.26Improved
replace MaskTheFace masks with our SparkAR masks in training (ablation)FaceNet · MFR2verification accuracy: 95.16 -> 96.22Improved
replace mask texture with black pixels inside mask (ablation)FaceNet · MFR2verification accuracy: 96.22 -> 93.04Hurt
replace mask texture with black pixels inside mask (ablation)VGG-face · MFR2verification accuracy: 91.26 -> 88.31Hurt

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