On Recognizing Occluded Faces in the Wild

Mustafa Ekrem Erakın; Ugur Demir; Hazım Kemal Ekenel

On Recognizing Occluded Faces in the Wild

Mustafa Ekrem Erakın, Ugur Demir, Hazım Kemal Ekenel

Faculty of Computer and Informatics

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

1 Citation (Scopus)

Abstract

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.

Original language	English
Title of host publication	BIOSIG 2021 - Proceedings of the 20th International Conference of the Biometrics Special Interest Group
Editors	Arslan Bromme, Christoph Busch, Naser Damer, Antitza Dantcheva, Marta Gomez-Barrero, Kiran Raja, Christian Rathgeb, Ana F. Sequeira, Andreas Uhl
Publisher	Gesellschaft fur Informatik (GI)
Pages	189-196
Number of pages	8
ISBN (Electronic)	9783885797098
Publication status	Published - 2021
Externally published	Yes
Event	20th International Conference of the Biometrics Special Interest Group, BIOSIG 2021 - Darmstadt, Germany Duration: 15 Sept 2021 → 17 Sept 2021

Publication series

Name	Lecture Notes in Informatics (LNI), Proceedings - Series of the Gesellschaft fur Informatik (GI)
Volume	P-315
ISSN (Print)	1617-5468

Conference

Conference	20th International Conference of the Biometrics Special Interest Group, BIOSIG 2021
Country/Territory	Germany
City	Darmstadt
Period	15/09/21 → 17/09/21

Bibliographical note

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

Keywords

Deep learning
Face occlusion
Face recognition
Real-world occluded faces

Cite this

Erakın, M. E., Demir, U., & Ekenel, H. K. (2021). On Recognizing Occluded Faces in the Wild. In A. Bromme, C. Busch, N. Damer, A. Dantcheva, M. Gomez-Barrero, K. Raja, C. Rathgeb, A. F. Sequeira, & A. Uhl (Eds.), BIOSIG 2021 - Proceedings of the 20th International Conference of the Biometrics Special Interest Group (pp. 189-196). (Lecture Notes in Informatics (LNI), Proceedings - Series of the Gesellschaft fur Informatik (GI); Vol. P-315). Gesellschaft fur Informatik (GI).

Erakın, Mustafa Ekrem ; Demir, Ugur ; Ekenel, Hazım Kemal. / On Recognizing Occluded Faces in the Wild. BIOSIG 2021 - Proceedings of the 20th International Conference of the Biometrics Special Interest Group. editor / Arslan Bromme ; Christoph Busch ; Naser Damer ; Antitza Dantcheva ; Marta Gomez-Barrero ; Kiran Raja ; Christian Rathgeb ; Ana F. Sequeira ; Andreas Uhl. Gesellschaft fur Informatik (GI), 2021. pp. 189-196 (Lecture Notes in Informatics (LNI), Proceedings - Series of the Gesellschaft fur Informatik (GI)).

@inproceedings{055d6b7f5705419bbf35486bc405a130,

title = "On Recognizing Occluded Faces in the Wild",

abstract = "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.",

keywords = "Deep learning, Face occlusion, Face recognition, Real-world occluded faces",

author = "Erakın, {Mustafa Ekrem} and Ugur Demir and Ekenel, {Hazım Kemal}",

note = "Publisher Copyright: {\textcopyright} 2021 Gesellschaft fur Informatik (GI). All rights reserved.; 20th International Conference of the Biometrics Special Interest Group, BIOSIG 2021 ; Conference date: 15-09-2021 Through 17-09-2021",

year = "2021",

language = "English",

series = "Lecture Notes in Informatics (LNI), Proceedings - Series of the Gesellschaft fur Informatik (GI)",

publisher = "Gesellschaft fur Informatik (GI)",

pages = "189--196",

editor = "Arslan Bromme and Christoph Busch and Naser Damer and Antitza Dantcheva and Marta Gomez-Barrero and Kiran Raja and Christian Rathgeb and Sequeira, {Ana F.} and Andreas Uhl",

booktitle = "BIOSIG 2021 - Proceedings of the 20th International Conference of the Biometrics Special Interest Group",

address = "Germany",

}

Erakın, ME, Demir, U & Ekenel, HK 2021, On Recognizing Occluded Faces in the Wild. in A Bromme, C Busch, N Damer, A Dantcheva, M Gomez-Barrero, K Raja, C Rathgeb, AF Sequeira & A Uhl (eds), BIOSIG 2021 - Proceedings of the 20th International Conference of the Biometrics Special Interest Group. Lecture Notes in Informatics (LNI), Proceedings - Series of the Gesellschaft fur Informatik (GI), vol. P-315, Gesellschaft fur Informatik (GI), pp. 189-196, 20th International Conference of the Biometrics Special Interest Group, BIOSIG 2021, Darmstadt, Germany, 15/09/21.

On Recognizing Occluded Faces in the Wild. / Erakın, Mustafa Ekrem; Demir, Ugur; Ekenel, Hazım Kemal.
BIOSIG 2021 - Proceedings of the 20th International Conference of the Biometrics Special Interest Group. ed. / Arslan Bromme; Christoph Busch; Naser Damer; Antitza Dantcheva; Marta Gomez-Barrero; Kiran Raja; Christian Rathgeb; Ana F. Sequeira; Andreas Uhl. Gesellschaft fur Informatik (GI), 2021. p. 189-196 (Lecture Notes in Informatics (LNI), Proceedings - Series of the Gesellschaft fur Informatik (GI); Vol. P-315).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - On Recognizing Occluded Faces in the Wild

AU - Erakın, Mustafa Ekrem

AU - Demir, Ugur

AU - Ekenel, Hazım Kemal

PY - 2021

Y1 - 2021

N2 - 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.

AB - 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.

KW - Deep learning

KW - Face occlusion

KW - Face recognition

KW - Real-world occluded faces

UR - http://www.scopus.com/inward/record.url?scp=85127456914&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:85127456914

T3 - Lecture Notes in Informatics (LNI), Proceedings - Series of the Gesellschaft fur Informatik (GI)

SP - 189

EP - 196

BT - BIOSIG 2021 - Proceedings of the 20th International Conference of the Biometrics Special Interest Group

A2 - Bromme, Arslan

A2 - Busch, Christoph

A2 - Damer, Naser

A2 - Dantcheva, Antitza

A2 - Gomez-Barrero, Marta

A2 - Raja, Kiran

A2 - Rathgeb, Christian

A2 - Sequeira, Ana F.

A2 - Uhl, Andreas

PB - Gesellschaft fur Informatik (GI)

T2 - 20th International Conference of the Biometrics Special Interest Group, BIOSIG 2021

Y2 - 15 September 2021 through 17 September 2021

ER -

Erakın ME, Demir U, Ekenel HK. On Recognizing Occluded Faces in the Wild. In Bromme A, Busch C, Damer N, Dantcheva A, Gomez-Barrero M, Raja K, Rathgeb C, Sequeira AF, Uhl A, editors, BIOSIG 2021 - Proceedings of the 20th International Conference of the Biometrics Special Interest Group. Gesellschaft fur Informatik (GI). 2021. p. 189-196. (Lecture Notes in Informatics (LNI), Proceedings - Series of the Gesellschaft fur Informatik (GI)).

On Recognizing Occluded Faces in the Wild

Abstract

Publication series

Conference

Bibliographical note

Keywords

Other files and links

Fingerprint

Cite this