Vision Transformer (VIT) Architecture for Robust Masked Face Recognition

The widespread adoption of facial masks during the COVID-19 pandemic significantly challenged existing facial recognition systems by occluding critical biometric features. This paper proposes a Vision Transformer (ViT) based approach for robust Masked Face Recognition (MFR). Unlike traditional Convolutional Neural Networks (CNNs) that rely on local receptive fields, the ViT architecture utilizes global self-attention to capture long-range dependencies, making it more resilient to the information loss caused by masks. We evaluate our approach on the MFR2 dataset, by implementing a standardized training methodology, and our model achieves a peak accuracy of 98.22%. This study demonstrates that transformer-based architectures, combined with specialized attention mechanisms and contrastive learning, offer a state-of-the-art solution for secure authentication in masked environments.

Attention Mechanisms, Masked Face Recognition, Vision Transformer.

1. Xu, "Based on the contrastive learning classifier for occluded face recognition," Procedia Computer Science, vol. 2025, 2025. DOI: 10.1016/j.procs.2025.08.148 [Google Scholar] [Crossref]

2. Zhu et al., "Joint holistic and masked face recognition," IEEE Transactions on Information Forensics and Security, 2023. DOI: 10.1109/TIFS.2023.3280717 [Google Scholar] [Crossref]

3. Zhao et al., "Masked Face Transformer," IEEE Transactions on Information Forensics and Security, 2023. DOI: 10.1109/tifs.2023.3322600 [Google Scholar] [Crossref]

4. "Joint Holistic and Masked Face Recognition," IEEE Transactions on Information Forensics and Security, 2023. DOI: 10.1109/tifs.2023.3280717 [Google Scholar] [Crossref]

5. Hosen et al., "HiMFR: A Hybrid Masked Face Recognition Through Face Inpainting," arXiv.org, 2022. DOI: 10.48550/arXiv.2209.08930 [Google Scholar] [Crossref]

6. Zhao et al., "Masked Face Transformer," IEEE Transactions on Information Forensics and Security, 2023. [Google Scholar] [Crossref]

7. "Robust Masked Face Recognition via Balanced Feature Matching," in Proc. 2022 IEEE International Conference on Consumer Electronics (ICCE), 2022. DOI: 10.1109/icce53296.2022.9730338 [Google Scholar] [Crossref]

8. Anwar et al., "Masked Face Recognition for Secure Authentication," arXiv: Computer Vision and Pattern Recognition, 2020. [Google Scholar] [Crossref]

9. "A Benchmark on Masked Face Recognition," in Proc. SIBGRAPI, 2022. DOI: 10.1109/sibgrapi55357.2022.9991785 [Google Scholar] [Crossref]

10. Iftikhar et al., "Masked Face Detection and Recognition Using a Unified Feature Extractor," in Proc. ICACS, 2024. DOI: 10.1109/icacs60934.2024.10473243 [Google Scholar] [Crossref]

11. "Ensemble Learning using Transformers and Convolutional Networks for Masked Face Recognition," arXiv.org, 2022. DOI: 10.48550/arxiv.2210.04816 [Google Scholar] [Crossref]

12. Mahmoud et al., "A Comprehensive Survey of Masked Faces: Recognition, Detection, and Unmasking," Applied Sciences, vol. 14, no. 19, 2024. DOI: 10.3390/app14198781 [Google Scholar] [Crossref]

13. "Towards Accurate and Lightweight Masked Face Recognition: An Experimental Evaluation," IEEE Access, vol. 2022, 2022. DOI: 10.1109/access.2021.3135255 [Google Scholar] [Crossref]

14. "A Survey on Computer Vision based Human Analysis in the COVID-19 Era," arXiv.org, 2022. DOI: 10.48550/arxiv.2211.03705 [Google Scholar] [Crossref]

• Investigating the Performance of Colour Oxide on the PS/CNSO Oil Paint Production
• Real-Time Fabric Defect Detection Using a Lightweight Deformable YOLO Network
• Recent Trends in the Stock Market: An Analytical Study of Market Dynamics, Investor Behaviour, and Technological Influence
• Evaluation of Structural Dynamics and Equilibrium State, with Case Study of Large Cantilever Projection for a 10- Storey Reinforced Concrete Building in Lagos, Nigeria.
• Multi-Criteria Evaluation of AI-Based Adaptive Learning Platforms in Global Higher Education: A Fuzzy AHP Perspective