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