《计算机应用研究》|Application Research of Computers

基于距离限定优化的人脸识别

Face recognition based on limit distance optimization

免费全文下载 (已被下载 次)  
获取PDF全文
作者 周胜阳,邹华,肖春霞
机构 武汉大学 计算机学院,武汉 430072
统计 摘要被查看 次,已被下载
文章编号 1001-3695(2019)03-059-0935-05
DOI 10.19734/j.issn.1001-3695.2017.10.1018
摘要 针对现有人脸识别方法对人脸角度、表情、姿态等因素较为敏感且准确率低的问题,提出了一种基于距离限定优化算法的人脸识别模型。该模型对人脸识别方法的改进有两点:a)利用LBP算子提取人脸图像纹理谱特征图,然后与原始人脸图像的R、G、B通道进行融合,将融合后的图像矩阵作为神经网络的输入,丰富了人脸的纹理特征;b)对误差函数进行改进,使用阈值和边界值约束特征向量的距离,对模型构建新的优化目标,使得相同对象的人脸图像在特征空间中具有较小的欧氏距离,不同对象的人脸图像在特征空间中具有较大的欧氏距离。通过在非限制场景下的LFW人脸库上进行实验,表明该模型准确率分别达到99.15%,能有效地提高人脸识别准确率,具有很好的鲁棒性。
关键词 人脸识别;特征提取;局部二值模式;二元误差函数;残差神经网络
基金项目 国家自然科学基金资助项目(61672390,61303026,61472288)
本文URL http://www.arocmag.com/article/01-2019-03-059.html
英文标题 Face recognition based on limit distance optimization
作者英文名 Zhou Shengyang, Zou Hua, Xiao Chunxia
机构英文名 SchoolofComputer,WuhanUniversity,Wuhan430072,China
英文摘要 Given the sensitivity of human face angle, expression and attitude as well as the low recognition precision, this paper presented a new face recognition algorithm based on distance optimization method.There were two revised points in the proposed method:a)The algorithm used the LBP operator to extract the texture map of the face image and then converged it with the R, G and B channels of the original image, then the neural network could take the fused image matrix as input, which enriched the human face texture features;b)In the processing of training, it reconstructed the loss function to improve the performance, made use of the thresold and margin to constrain the distance of the feature vector, and built a new optimization target for the model, which could limit the faces of the same person have small euclidean distances and faces of distinct people have large distances.Experiments on the unconstrained scenes of the LFW face database show that the accuracy of the model is 99.15% respectively, indict the model can improve the accuracy of face recognition with strong robustness.
英文关键词 face recognition; feature extraction; local binary pattern; binary loss function; residual network
参考文献 查看稿件参考文献
  [1] Sun Yi, Wang Xiaogang, Tang Xiaoou. Hybrid deep learning for face verification[C] //Proc of IEEE International Conference on Computer Vision. Piscataway, NJ:IEEE Press, 2013:1489-1496.
[2] Taigman Y, Yang M, Ranzato M A, et al. DeepFace:closing the gap to human-level performance in face verification[C] //Proc of IEEE Conference on Computer Vision and Pattern Recognition. Piscataway, NJ:IEEE Press, 2014:1701-1708.
[3] Schroff F, Kalenichenko D, Philbin J. FaceNet:a unified embedding for face recognition and clustering[C] //Proc of IEEE Conference on Computer Vision and Pattern Recognition. Piscataway, NJ:IEEE Press, 2015:815-823.
[4] Bartlett M S, Movellan J R, Sejnowski T J. Face recognition by independent component analysis[J] . IEEE Trans on Neural Networks, 2002, 13(6):1450-1464.
[5] Albiol A, Monzo D, Martin A, et al. Face recognition using HOG-EBGM[J] . Pattern Recognition Letters, 2008, 29(10):1537-1543.
[6] 万源, 李欢欢, 吴克风, 等. LBP和HOG的分层特征融合的人脸识别[J] . 计算机辅助设计与图形学学报, 2015, 27(4):640-650. (Wan Yuan, Li Huanhuan, Wu Kefeng, et al. Hierarchical feature fusion of LBP and HOG[J] . Journal of Computer Aided Design and Graphics, 2015, 27(4):640-650. )
[7] 徐梦珂, 许道云, 魏明俊. 基于 2D-KPCA 的拉普拉斯特征映射人脸识别[J] . 计算机应用研究, 2017, 34(7):2212-2215, 2220. (Xu Mengke, Xu Daoyun, Wei Mingjun. Laplacian feature mapping face recognition based on 2D-KPCA[J] . Application Research of Computers, 2017, 34(7):2212-2215, 2220. )
[8] Ahonen T, Hadid A, Pietikinen M. Face recognition with local binary patterns[C] //Proc of European Conference on Computer Vision. Berlin:Springer, 2004:469-481.
[9] Simonyan K, Vedaldi A, Zisserman A. Learning local feature descriptors using convex optimisation[J] . IEEE Trans on Pattern Analysis and Machine Intelligence, 2014, 36(8):1573-1585.
[10] 杨赛, 赵春霞, 刘凡, 等. 一种基于多种特征融合的人脸识别算法[J] . 计算机辅助设计与图形学学报, 2017, 29(9):1667-1672. (Yang Sai, Zhao Chunxia, Liu Fan, et al. A face recognition algorithm based on multi-feature fusion[J] . Journal of Computer Aided Design and Graphics, 2017, 29(9):1667-1672. )
[11] Zhang Ling, Zhang Qing, Xiao Chunxia. Shadow remover:image shadow removal based on illumination recovering optimization[J] . IEEE Trans on Image Processing, 2015, 24(11):4623-4636.
[12] Xiao Chunxia, Gan Jiajia. Fast image dehazing using guided joint bilateral filter[J] . The Visual Computer, 2012, 28(6-8):713-721.
[13] Xiao Chunxia, Nie Yongwei, Hua Wei, et al. Fast multi-scale joint bilateral texture upsampling[J] . The Visual Computer, 2010, 26(4):263-275.
[14] Sun Yi, Wang Xiaogang, Tang Xiaoou. Deep learning face representation from predicting 10, 000 classes[C] //Proc of IEEE Conference on Computer Vision and Pattern Recognition. Piscataway, NJ:IEEE Press, 2014:1891-1898.
[15] Sun Yi, Wang Xiaogang, Tang Xiaoou, et al. Deep learning face representation by joint identification-verification[C] //Advances in Neural Information Processing Systems. 2014:1988-1996.
[16] Xi Meng, Chen Liang, Polajnar D, et al. Local binary pattern network:a deep learning approach for face recognition[C] //Proc of IEEE International Conference on Image Processing. Piscataway, NJ:IEEE Press, 2016:3224-3228.
[17] He Kaiming, Zhang Xiangyu, Ren Shaoqing, et al. Deep residual learning for image recognition[C] //Proc of IEEE Conference on Computer Vision and Pattern Recognition. Piscataway, NJ:IEEE Press, 2016:770-778.
[18] Heikkil M, Pietikinen M, Schmid C. Description of interest regions with local binary patterns[J] . Pattern Recognition, 2009, 42(3):425-436.
[19] Ng H W, Winkler S. A data-driven approach to cleaning large face datasets[C] //Proc of IEEE International Conference on Image Processing. Piscataway, NJ:IEEE Press, 2014:343-347.
[20] Huang G B, Mattar M, Berg T, et al. Labeled faces in the wild:a database for studying face recognition in unconstrained environments[D] . Amherst:University of Massachusetts, 2007.
收稿日期 2017/10/22
修回日期 2017/12/18
页码 935-939
中图分类号 TP391.41
文献标志码 A