Balanced Masked and Standard Face Recognition

Qi, Delong; Hu, Kangli; Tan, Weijun; Yao, Qi; Liu, Jingfeng

doi:10.1109/iccvw54120.2021.00174

Cited by 6 publications

(4 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…By continually recording the local and global information of the same facial area, ResSaNet can provide impressive performance of more than 78% TAR on both masked and non-masked testing data. [36] To fast decrease the input resolution, a new stem unit drop block [37] was introduced. This unit contributed to the improvement of the ResNet backbone, which is helpful for effective MFR feature extraction.…”

Section: Mask Robust Methodsmentioning

confidence: 99%

A Survey on Masked Face Recognition Amid the COVID-19 Pandemic

Zhong

2023

Frontiers in Artificial Intelligence and Applications

View full text Add to dashboard Cite

Against the backdrop of the COVID-19 pandemic, the ubiquitous adoption of masks has caused significant hurdles to conventional face recognition technologies and their applications. In order to address these problems, a great deal of research related to masked face recognition has been put forward. Although masked face recognition has made great progress, the review or summary about this field is scanty. Therefore, this paper proposed a survey on masked face recognition. Firstly, the datasets of MFR and the mainstream deep learning methods are introduced. In addition, some common evaluation metrics are explained. Finally, this paper discusses the potential challenges of MFR and make conclusion.

show abstract

Section: Mask Robust Methodsmentioning

confidence: 99%

A Survey on Masked Face Recognition Amid the COVID-19 Pandemic

Zhong

2023

Frontiers in Artificial Intelligence and Applications

View full text Add to dashboard Cite

show abstract

“…It's known that occlusion is one of the factors affecting the effect of face recognition 30,31 . Standard face recognition methods are good at extracting features from common faces, but it does not perform well on faces with occlusions 32 . The loss of facial feature caused by occlusion makes it difficult to learn discriminative feature by standard face recognition models.…”

Section: Related Workmentioning

confidence: 99%

“…Unlike the task of occluded face recognition, the masked face recognition task is more complex. The location of masks always covers a half or more area of faces, which causes more feature missing 32 . The other challenge for masked face recognition is the lack of high-quality large-scale datasets.…”

Section: Masked Face Recognitionmentioning

confidence: 99%

A joint feature aggregation method for robust masked face recognition

Zhu²,

Bai³

2022

Fourteenth International Conference on Digital Image Processing (ICDIP 2022)

View full text Add to dashboard Cite

Masked face recognition becomes an important issue of prevention and monitor in outbreak of COVID-19. Due to loss of facial features caused by masks, unmasked face recognition could not identify the specific person well. Current masked faces methods focus on local features from the unmasked regions or recover masked faces to fit standard face recognition models. These methods only focus on partial information of faces thus these features are not robust enough to deal with complex situations. To solve this problem, we propose a joint feature aggregation method for robust masked face recognition. Firstly, we design a multi-module feature extraction network to extract different features, including local module (LM), global module (GM), and recovery module (RM). Our method not only extracts global features from the original masked faces but also extracts local features from the unmasked area since it is a discriminative part of masked faces. Specially, we utilize a pretrained recovery model to recover masked faces and get some recovery features from the recovered faces. Finally, features from three modules are aggregated as a joint feature of masked faces. The joint feature enhances the feature representation of masked faces thus it is more discriminative and robust than that in previous methods. Experiments show that our method can achieve better performance than previous methods on LFW dataset.

show abstract

“…Above all, the robust feature extraction methods all have limited attention performance and unsatisfactory accuracy on mask-wearing occlusion tasks. On the other hand, most of the occluded region recovery methods require additional information of occluded regions, which becomes challenging when dealing with large occlusion in large-scale masked face recognition benchmarks [15,16]. Mask-wearing occlusion is still a big challenge in face recognition, especially in real-world uncooperative situations.…”

Section: Introductionmentioning

confidence: 99%

Attention-Mechanism-Based Models for Unconstrained Face Recognition with Mask Occlusion

Zhang,

Zhang

2023

Electronics

View full text Add to dashboard Cite

Masks cover most areas of the face, resulting in a serious loss of facial identity information; thus, how to alleviate or eliminate the negative impact of occlusion is a significant problem in the field of unconstrained face recognition. Inspired by the successful application of attention mechanisms and capsule networks in computer vision, we propose ECA-Inception-Resnet-Caps, which is a novel framework based on Inception-Resnet-v1 for learning discriminative face features in unconstrained mask-wearing conditions. Firstly, Squeeze-and-Excitation (SE) modules and Efficient Channel Attention (ECA) modules are applied to Inception-Resnet-v1 to increase the attention on unoccluded face areas, which is used to eliminate the negative impact of occlusion during feature extraction. Secondly, the effects of the two attention mechanisms on the different modules in Inception-Resnet-v1 are compared and analyzed, which is the foundation for further constructing the ECA-Inception-Resnet-Caps framework. Finally, ECA-Inception-Resnet-Caps is obtained by improving Inception-Resnet-v1 with capsule modules, which is explored to increase the interpretability and generalization of the model after reducing the negative impact of occlusion. The experimental results demonstrate that both attention mechanisms and the capsule network can effectively enhance the performance of Inception-Resnet-v1 for face recognition in occlusion tasks, with the ECA-Inception-Resnet-Caps model being the most effective, achieving an accuracy of 94.32%, which is 1.42% better than the baseline model.

show abstract

Balanced Masked and Standard Face Recognition

Cited by 6 publications

References 19 publications

A Survey on Masked Face Recognition Amid the COVID-19 Pandemic

A Survey on Masked Face Recognition Amid the COVID-19 Pandemic

A joint feature aggregation method for robust masked face recognition

Attention-Mechanism-Based Models for Unconstrained Face Recognition with Mask Occlusion

Contact Info

Product

Resources

About