MLFW: A Database for Face Recognition on Masked Faces

13 Sep 2021  ·  Chengrui Wang, Han Fang, Yaoyao Zhong, Weihong Deng ·

As more and more people begin to wear masks due to current COVID-19 pandemic, existing face recognition systems may encounter severe performance degradation when recognizing masked faces. To figure out the impact of masks on face recognition model, we build a simple but effective tool to generate masked faces from unmasked faces automatically, and construct a new database called Masked LFW (MLFW) based on Cross-Age LFW (CALFW) database. The mask on the masked face generated by our method has good visual consistency with the original face. Moreover, we collect various mask templates, covering most of the common styles appeared in the daily life, to achieve diverse generation effects. Considering realistic scenarios, we design three kinds of combinations of face pairs. The recognition accuracy of SOTA models declines 5%-16% on MLFW database compared with the accuracy on the original images. MLFW database can be viewed and downloaded at \url{}.

PDF Abstract
No code implementations yet. Submit your code now


Introduced in the Paper:


Used in the Paper:

VGGFace2 CASIA-WebFace

Results from the Paper

Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Face Recognition MLFW MS1MV2, R100, SFace Accuracy 91.57 # 1
Face Recognition MLFW MS1MV2, R100, Curricularface Accuracy 90.43 # 3
Face Recognition MLFW MS1MV2, R100, Arcface Accuracy 90.57 # 2
Face Recognition MLFW VGGFace2, R50, ArcFace Accuracy 85.95 # 4
Face Recognition MLFW CASIA-WebFace, R50, CosFace Accuracy 82.52 # 5
Face Recognition MLFW Private-Asia, R50, ArcFace Accuracy 77.20 # 6


No methods listed for this paper. Add relevant methods here