Consistent Sub-Decision Network for Low-Quality Masked Face Recognition
IEEE Signal Processing Letters
; 29:1147-1151, 2022.
Article
in English
| ProQuest Central | ID: covidwho-1840273
ABSTRACT
The COVID-19 pandemic makes wearing masks mandatory in supermarkets, pharmacies, public transport, etc. Existing facial recognition systems encounter severe performance degradation as the masks occlude key facial regions. Recently, simulation-based methods are proposed to generate masked faces from unmasked faces. However, among simulated faces, there are low-quality samples with negative occlusion, which leads to ambiguous or absent facial features. In this paper, we propose a consistent sub-decision network to obtain sub-decisions that correspond to different facial regions and constrain sub-decisions by weighted bidirectional KL divergence to make the network concentrate on the upper faces without occlusion. In addition, we perform knowledge distillation to drive the masked face embeddings towards an approximation of the original data distribution to mitigate the information loss. Experiments show that the proposed method performs better than the baseline on public masked face recognition datasets, i.e., RMFD, MFR2, and MLFW.
Engineering--Electrical Engineering; Face recognition; Faces; Feature extraction; Facial features; Training; Knowledge engineering; Drives; COVID-19; masked face recognition; low-quality; Distillation; Masks; Performance degradation; Facial recognition technology; Public transportation; Occlusion; Decisions
Full text:
Available
Collection:
Databases of international organizations
Database:
ProQuest Central
Language:
English
Journal:
IEEE Signal Processing Letters
Year:
2022
Document Type:
Article
Similar
MEDLINE
...
LILACS
LIS