Mask occlusion face recognition algorithm based on neural network model with optimized loss function
-
摘要: 矿井人脸信息识别是矿井人员管理、安全生产的重要环节和保障措施，但矿井的特殊环境可能会出现面部遮挡或不清等情况，尤其是在煤矿井下多粉尘等区域，面罩往往被普遍使用，这些都会增加人脸识别的难度。为了能够对面罩遮挡状态下的人脸图像进行准确识别，提出了一种基于损失函数优化神经网络模型的面罩遮挡人脸识别算法。通过基于人脸关键点检测的面罩遮挡人脸生成算法将常规人脸数据集扩充为面罩遮挡人脸数据集，缓解煤矿面罩遮挡人脸数据不足的问题。根据面罩遮挡人脸图像的特点使用损失函数进行模型训练，使用更优的损失函数代替三元组损失，同时添加注意力机制使模型更加关注于未被遮挡的区域，使模型能更好地提取面罩遮挡状态下的人脸特征。通过大量实验证明，基于损失函数优化神经网络模型的面罩遮挡人脸识别算法不仅在提取人脸特征时更加关注于未被面罩遮挡的上半部分区域，而且在面对噪声时有较强的鲁棒性，在MFR2数据集中采用该算法将原始网络的准确率由84.3%提高到了98.5%,相较于其他方法具有较高的识别准确率，能够在面罩遮挡状态下完成人脸识别任务。
-
-
[1] 程德强, 寇旗旗, 江鹤, 等.全矿井智能视频分析关键技术综述[J].工矿自动化, 2023, 49(11):1-21. [2] 王国法, 杜毅博, 任怀伟, 等.智能化煤矿顶层设计研究与实践[J].煤炭学报, 2020, 45(6):1909-1924. [3] DENG Jiankang, GUO Jia, YANG Jing, et al.ArcFace:additive angular margin loss for deep face recognition[Z/OL].[2023-12-10]. https://doi.org/10.48550/arXiv.1801.07698.
[4] WANG Hao, WANG Yitong, ZHOU Zheng, et al.CosFace:large margin cosine loss for deep face recognition[C].IEEE/CVF Conference on Computer Vision and Pattern Recognition, Salt Lake City, 2018.DOI: 10.1109/CVPR.2018.00552.
[5] HUANG G B, MATTAR M, BERG T, et al.Labeled faces in the wild:a database for studying face recognition in unconstrained environments[C].Workshop on Faces in 'Real-Life' Images:Detection, Alignment, and Recognition, Piscatawa, 2008:1-8.
[6] JIA Hongjun, MARTINEZ A M.Support vector machines in face recognition with occlusions[C].IEEE Conference on Computer Vision and Pattern Recognition, Miami, 2009.DOI: 10.1109/CVPR.2009.5206862.
[7] WU C Y, DING J J.Occluded face recognition using low-rank regression with generalized gradient direction[Z/OL].[2023-12-17]. http://arxiv.org/abs/1906.02429.
[8] 朱孟刚, 郑广海.遮挡人脸识别算法改进方法综述[J].计算机科学与应用, 2022, 12(6):1569-1579. [9] MONTERO D, NIETO M, AGINAKO N, et al.Boosting masked face recognition with multi-task arcface[C].The 16th International Conference on Signal-Image Technology & Internet-Based Systems, Dijon, 2022.DOI: 10.1109/SITIS57111.2022.00042.
[10] FENG Tao, XU Liangpeng, YUAN Hangjie, et al.Towards mask-robust face recognition[C].IEEE/CVF International Conference on Computer Vision Workshops, Montreal, 2021.DOI: 10.1109/ICCVW54120.2021.00173.
[11] BOUTROS F, DAMER N, KIRCHBUCHNER F, et al.Self-restrained triplet loss for accurate masked face recognition[Z/OL].[2023-12-17]. https://arxiv.org/abs/2103.01716v2.
[12] DING Feifei, PENG Peixi, HUANG Yangru, et al.Masked face recognition with latent part detection[C].The 28th ACM International Conference on Multimedia, New York, 2020.DOI: 10.1145/3394171.3413731.
[13] HE Lingxiao, LI Haiqing, ZHANG Qi, et al.Dynamic feature learning for partial face recognition[C].IEEE/CVF Conference on Computer Vision and Pattern Recognition, Salt Lake City, 2018.DOI: 10.1109/CVPR.2018.00737.
[14] CAI Jiancheng, HAN Hu, CUI Jiyun.et al.Semi-supervised natural face de-occlusion[C].IEEE Transactions on Information Forensics and Security, 2020.DOI: 10.1109/TIFS.2020.3023793.
[15] GENG Mengyue, PENG Peixi, HUANG Yangru, et al.Masked face recognition with generative data augmentation and domain constrained ranking[C].Medical Image Computing and Computer Assisted Intervention, 2020.
[16] 王国法.煤矿智能化最新技术进展与问题探讨[J].煤炭科学技术, 2022, 50(1):1-27. [17] 刘晓阳, 霍祎炜.多尺度分区统一化LBP算子井下人员人脸识别方法[J].煤炭科学技术, 2019, 47(12):116-123. [18] 陈岸明, 林群雄, 刘伟强.基于对比学习的多特征融合戴口罩人脸识别[J].计算机应用研究, 2024, 41(1):277-281, 287. [19] 王均利, 李佳悦, 李秉天, 等.基于深度学习的煤矿井下低光照人脸检测方法[J].工矿自动化, 2023, 49(11):145-150. [20] DENG Jiankang, GUO Jia, VERVERAS E, et al.RetinaFace:single-shot multi-level face localization in the wild [C].IEEE/CVF Conference on Computer Vision and Pattern Recognition, Seattle, 2020.DOI: 10.1109/CVPR42600.2020.00525.
[21] SCHROFF F, KALENICHENKO D, PHILBIN J.FaceNet:A uni-fied embedding for face recognition and clustering[C].IEEE Conference on Computer Vision and Pattern Recognition, Boston, 2015.DOI: 10.1109/CVPR.2015.7298682.
[22] WANG Hao, WANG Yitong, ZHOU Zheng, et al.CosFace:large margin cosine loss for deep face recognition[C].IEEE/CVF Conference on Computer Vision and Pattern Recognition, Salt Lake City, 2018.DOI: 10.1109/CVPR.2018.00552.
[23] DENG Jiankang, GUO Jia, ZAFEIRIOU S.ArcFace:additive angular margin loss for deep face recognition[Z/OL].[2023-12-20]. https://arxiv.org/abs/1801.07698.
计量
- 文章访问数: 9
- HTML全文浏览量: 0
- PDF下载量: 2
下载:
