Abstract
Face masks bring a new challenge to face recognition systems especially against the background of the COVID-19 pandemic. In this paper, a method used for mitigating the negative effects of mask defects on face recognition is proposed. Firstly, a low-cost, accurate method of masked face synthesis, i.e. mask transfer, is proposed for data augmentation. Secondly, an attention-aware masked face recognition (AMaskNet) is proposed to improve the performance of masked face recognition, which includes two modules: a feature extractor and a contribution estimator. Therein, the contribution estimator is employed to learn the contribution of the feature elements, thus achieving refined feature representation by simple matrix multiplications. Meanwhile, the end-to-end training strategy is utilized to optimize the entire model. Finally, a mask-aware similarity matching strategy (MS) is adopted to improve the performance in the inference stage. The experiments show that the proposed method consistently outperforms on three masked face recognition datasets: RMFRD, COX and Public-IvS. Meanwhile, a qualitative analysis experiments using CAM indicate that the contribution learned by AMaskNet is more conducive to masked face recognition.
Highlights
The COVID-19 pandemic has created a global disaster: the World Health Organization (WHO) and Centers for Disease Control and Prevention (CDC) have suggested everyone should wear a mask in a public setting especially when other social distancing measures are difficult to maintain [4]
The comparison between R34-AMaskNet and R34-Mask shows that AMaskNet is able to improve the performance, especially for masked face recognition, e.g. a 1.1 percentage points improvement in the case of W/W on the Cam2 of COX, which indicates that the proposed contribution estimator can learn an effective contribution matrix and automatically assign higher weights to the feature map activated by the non-masked facial parts and lower weights to those that are activated by masked facial parts
For R34Mask and R34-AMaskNet, the data augmentation strategy has been adopted, it is still improved in most cases, e.g., a 0.7 percentage points improvement for the R34AMaskNet on the Cam2 of COX between images treated without mask-transfer to those with mask-transfer
Summary
The COVID-19 pandemic has created a global disaster: the World Health Organization (WHO) and Centers for Disease Control and Prevention (CDC) have suggested everyone should wear a mask in a public setting especially when other social distancing measures are difficult to maintain [4]. It is essential to study the effect of wearing face masks on the behavior of face recognition systems and design mitigation techniques to offset the inevitable performance loss. The use of face masks triggers a significant research challenge: firstly, it is necessary to collect a large-scale training dataset, which includes the different types of faces with masks. To collect such a large-scale training dataset, on the one hand, it is time consuming and incurs higher labor cost, on the other hand, maintaining the diversity of data in such datasets is a slow process. It is necessary to mitigate the performance loss from the perspective of model design according to the characteristics of face masks
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
