Masked Face Transformer
文献类型:期刊论文
作者 | Zhao, Weisong1,2; Zhu, Xiangyu3,4; Guo, Kaiwen3,4; Shi, Haichao1,2; Zhang, Xiao-Yu1,2; Lei, Zhen3,4,5 |
刊名 | IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY |
出版日期 | 2024 |
卷号 | 19页码:265-279 |
ISSN号 | 1556-6013 |
关键词 | Face recognition Transformers Feature extraction Training Task analysis Costs COVID-19 Masked face recognition face recognition transformer |
DOI | 10.1109/TIFS.2023.3322600 |
通讯作者 | Lei, Zhen(zlei@nlpr.ia.ac.cn) |
英文摘要 | The COVID-19 pandemic makes wearing masks mandatory. Existing CNN-based face recognition (FR) systems suffer from severe performance degradation as masks occlude the vital facial regions. Recently, Vision Transformers have shown promising performance in various vision tasks with quadratic computation costs. Swin Transformer first proposes a successive window attention mechanism allowing the cross-window connection and more computational efficiency. Despite its potential, the deployment of Swin Transformer in masked face recognition encounters two challenges: 1) the attention range is insufficient to capture locally compatible face regions. 2) Masked face recognition can be defined as an occlusion-robust classification task with a known occlusion position, i.e., the position of the mask is minor-varying, which is overlooked but efficient in improving the model's recognition accuracy. To alleviate the above problem, we propose a Masked Face Transformer (MFT) with Masked Face-compatible Attention (MFA). The proposed MFA 1) introduces two additional window partition configurations, e.g., row shift and column shift, to enlarge the attention range in Swin with invariant computation costs, and 2) suppresses the interaction between the masked and non-masked regions to retain their discrepancies. Additionally, as mask occlusion leads to a separation between the masked and non-masked samples of the same identity, we propose to explore the relationship between them by a ClassFormer module to enhance intra-class aggregation. Extensive experiments show that MFT outperforms state-of-the-art masked face recognition methods in both simulated and real masked face testing datasets. |
WOS关键词 | RECOGNITION ; ROBUST |
资助项目 | Chinese National Natural Science Foundation Projects |
WOS研究方向 | Computer Science ; Engineering |
语种 | 英语 |
出版者 | IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC |
WOS记录号 | WOS:001123966000006 |
资助机构 | Chinese National Natural Science Foundation Projects |
源URL | [http://ir.ia.ac.cn/handle/173211/54943] |
专题 | 多模态人工智能系统全国重点实验室 |
通讯作者 | Lei, Zhen |
作者单位 | 1.Chinese Acad Sci, Inst Informat Engn, Beijing 100190, Peoples R China 2.Univ Chinese Acad Sci, Sch Cyber Secur, Beijing 101408, Peoples R China 3.Chinese Acad Sci, Inst Automat, State Key Lab Multimodal Artificial Intelligence S, Beijing 100190, Peoples R China 4.Univ Chinese Acad Sci UCAS, Sch Artificial Intelligence, Beijing 101408, Peoples R China 5.Chinese Acad Sci, Hong Kong Inst Sci & Innovat, CAIR, Hong Kong, Peoples R China |
推荐引用方式 GB/T 7714 | Zhao, Weisong,Zhu, Xiangyu,Guo, Kaiwen,et al. Masked Face Transformer[J]. IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY,2024,19:265-279. |
APA | Zhao, Weisong,Zhu, Xiangyu,Guo, Kaiwen,Shi, Haichao,Zhang, Xiao-Yu,&Lei, Zhen.(2024).Masked Face Transformer.IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY,19,265-279. |
MLA | Zhao, Weisong,et al."Masked Face Transformer".IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY 19(2024):265-279. |
入库方式: OAI收割
来源:自动化研究所
浏览0
下载0
收藏0
其他版本
除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。