Ana gezinime geç Aramaya geç Ana içeriğe geç

On Recognizing Occluded Faces in the Wild

Araştırma sonucu: Kitap/Rapor/Konferans Bildirisinde BölümKonferans katkısıbilirkişi

1 Atıf (Scopus)

Özet

Facial appearance variations due to occlusion has been one of the main challenges for face recognition systems. To facilitate further research in this area, it is necessary and important to have occluded face datasets collected from real-world, as synthetically generated occluded faces cannot represent the nature of the problem. In this paper, we present the Real World Occluded Faces (ROF) dataset, that contains faces with both upper face occlusion, due to sunglasses, and lower face occlusion, due to masks. We propose two evaluation protocols for this dataset. Benchmark experiments on the dataset have shown that no matter how powerful the deep face representation models are, their performance degrades significantly when they are tested on real-world occluded faces. It is observed that the performance drop is far less when the models are tested on synthetically generated occluded faces. The ROF dataset and the associated evaluation protocols are publicly available at the following link https://github.com/ekremerakin/RealWorldOccludedFaces.

Orijinal dilİngilizce
Ana bilgisayar yayını başlığıBIOSIG 2021 - Proceedings of the 20th International Conference of the Biometrics Special Interest Group
EditörlerArslan Bromme, Christoph Busch, Naser Damer, Antitza Dantcheva, Marta Gomez-Barrero, Kiran Raja, Christian Rathgeb, Ana F. Sequeira, Andreas Uhl
YayınlayanGesellschaft fur Informatik (GI)
Sayfalar189-196
Sayfa sayısı8
ISBN (Elektronik)9783885797098
Yayın durumuYayınlandı - 2021
Harici olarak yayınlandıEvet
Etkinlik20th International Conference of the Biometrics Special Interest Group, BIOSIG 2021 - Darmstadt, Germany
Süre: 15 Eyl 202117 Eyl 2021

Yayın serisi

AdıLecture Notes in Informatics (LNI), Proceedings - Series of the Gesellschaft fur Informatik (GI)
HacimP-315
ISSN (Basılı)1617-5468
ISSN (Elektronik)2944-7682

???event.eventtypes.event.conference???

???event.eventtypes.event.conference???20th International Conference of the Biometrics Special Interest Group, BIOSIG 2021
Ülke/BölgeGermany
ŞehirDarmstadt
Periyot15/09/2117/09/21

Bibliyografik not

Publisher Copyright:
© 2021 Gesellschaft fur Informatik (GI). All rights reserved.

Finansman

This study is supported by the Istanbul Technical University Research Fund, ITU BAP, project no. 42547 and by the Scientific and Technological Research Council of Turkey (TUBITAK) project no. 120N011.

FinansörlerFinansör numarası
ITU BAP42547
Istanbul Technical University Research Fund
TUBITAK120N011
Türkiye Bilimsel ve Teknolojik Araştirma Kurumu

    Parmak izi

    On Recognizing Occluded Faces in the Wild' araştırma başlıklarına git. Birlikte benzersiz bir parmak izi oluştururlar.

    Alıntı Yap